From 40430081b2bd0da9252ae2b19da7324d5344170f Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 08:59:13 +1000
Subject: [PATCH 01/28] Add xAI API key option to configuration
---
lib/ruby_llm/configuration.rb | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/ruby_llm/configuration.rb b/lib/ruby_llm/configuration.rb
index 377c6afc5..0701a442d 100644
--- a/lib/ruby_llm/configuration.rb
+++ b/lib/ruby_llm/configuration.rb
@@ -12,6 +12,7 @@ class Configuration
:gemini_api_key,
:deepseek_api_key,
:perplexity_api_key,
+ :xai_api_key,
:bedrock_api_key,
:bedrock_secret_key,
:bedrock_region,
From 120f9b19dd9456fa426addefe9cc9a1a981de7e6 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:00:37 +1000
Subject: [PATCH 02/28] Add initial xAI Provider
---
lib/ruby_llm/providers/xai.rb | 40 ++++++++++++++++++++++++++++++
lib/ruby_llm/providers/xai/chat.rb | 15 +++++++++++
2 files changed, 55 insertions(+)
create mode 100644 lib/ruby_llm/providers/xai.rb
create mode 100644 lib/ruby_llm/providers/xai/chat.rb
diff --git a/lib/ruby_llm/providers/xai.rb b/lib/ruby_llm/providers/xai.rb
new file mode 100644
index 000000000..b4a832eae
--- /dev/null
+++ b/lib/ruby_llm/providers/xai.rb
@@ -0,0 +1,40 @@
+# frozen_string_literal: true
+
+module RubyLLM
+ module Providers
+ # xAI API integration
+ class XAI < OpenAI
+ include XAI::Chat
+
+ def api_base
+ 'https://api.x.ai/v1'
+ end
+
+ def headers
+ {
+ 'Authorization' => "Bearer #{@config.xai_api_key}",
+ 'Content-Type' => 'application/json'
+ }
+ end
+
+ # xAI uses a different error format than OpenAI
+ # {"code": "...", "error": "..."}
+ def parse_error(response)
+ return if response.body.empty?
+
+ body = try_parse_json(response.body)
+ case body
+ when Hash then body['error']
+ when Array then body.map { |part| part['error'] }.join('. ')
+ else body
+ end
+ end
+
+ class << self
+ def configuration_requirements
+ %i[xai_api_key]
+ end
+ end
+ end
+ end
+end
diff --git a/lib/ruby_llm/providers/xai/chat.rb b/lib/ruby_llm/providers/xai/chat.rb
new file mode 100644
index 000000000..9fb2c9af5
--- /dev/null
+++ b/lib/ruby_llm/providers/xai/chat.rb
@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+module RubyLLM
+ module Providers
+ class XAI
+ # Chat implementation for xAI
+ # https://docs.x.ai/docs/api-reference#chat-completions
+ module Chat
+ def format_role(role)
+ role.to_s
+ end
+ end
+ end
+ end
+end
From 29bc4918c558694ccce1c099dfc2303b962bc5af Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:00:44 +1000
Subject: [PATCH 03/28] Registre xAI as Provider
---
lib/ruby_llm.rb | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/lib/ruby_llm.rb b/lib/ruby_llm.rb
index 0c4d0689a..afb29a61c 100644
--- a/lib/ruby_llm.rb
+++ b/lib/ruby_llm.rb
@@ -21,7 +21,8 @@
'openrouter' => 'OpenRouter',
'gpustack' => 'GPUStack',
'mistral' => 'Mistral',
- 'pdf' => 'PDF'
+ 'pdf' => 'PDF',
+ 'xai' => 'XAI'
)
loader.ignore("#{__dir__}/tasks")
loader.ignore("#{__dir__}/ruby_llm/railtie")
@@ -88,6 +89,7 @@ def logger
RubyLLM::Provider.register :openai, RubyLLM::Providers::OpenAI
RubyLLM::Provider.register :openrouter, RubyLLM::Providers::OpenRouter
RubyLLM::Provider.register :perplexity, RubyLLM::Providers::Perplexity
+RubyLLM::Provider.register :xai, RubyLLM::Providers::XAI
if defined?(Rails::Railtie)
require 'ruby_llm/railtie'
From 93e1bd777b45da783e367fc21735cc57a495d737 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:00:57 +1000
Subject: [PATCH 04/28] Add xAI Capabilities
---
lib/ruby_llm/providers/xai.rb | 5 +
lib/ruby_llm/providers/xai/capabilities.rb | 166 +++++++++++++++++++++
2 files changed, 171 insertions(+)
create mode 100644 lib/ruby_llm/providers/xai/capabilities.rb
diff --git a/lib/ruby_llm/providers/xai.rb b/lib/ruby_llm/providers/xai.rb
index b4a832eae..3c157117d 100644
--- a/lib/ruby_llm/providers/xai.rb
+++ b/lib/ruby_llm/providers/xai.rb
@@ -4,6 +4,7 @@ module RubyLLM
module Providers
# xAI API integration
class XAI < OpenAI
+ include XAI::Capabilities
include XAI::Chat
def api_base
@@ -31,6 +32,10 @@ def parse_error(response)
end
class << self
+ def capabilities
+ XAI::Capabilities
+ end
+
def configuration_requirements
%i[xai_api_key]
end
diff --git a/lib/ruby_llm/providers/xai/capabilities.rb b/lib/ruby_llm/providers/xai/capabilities.rb
new file mode 100644
index 000000000..a08ccb333
--- /dev/null
+++ b/lib/ruby_llm/providers/xai/capabilities.rb
@@ -0,0 +1,166 @@
+# frozen_string_literal: true
+
+module RubyLLM
+ module Providers
+ class XAI
+ # Determines capabilities and pricing for xAI (Grok) models
+ # - https://docs.x.ai/docs/models
+ module Capabilities
+ module_function
+
+ # rubocop:disable Naming/VariableNumber
+ MODEL_PATTERNS = {
+ grok_2: /^grok-2(?!-vision)/,
+ grok_2_vision: /^grok-2-vision/,
+ grok_2_image: /^grok-2-image/,
+ grok_3: /^grok-3(?!-(?:fast|mini))/,
+ grok_3_fast: /^grok-3-fast/,
+ grok_3_mini: /^grok-3-mini(?!-fast)/,
+ grok_3_mini_fast: /^grok-3-mini-fast/,
+ grok_4: /^grok-4/
+ }.freeze
+ # rubocop:enable Naming/VariableNumber
+
+ def context_window_for(model_id)
+ case model_family(model_id)
+ when 'grok_4' then 256_000
+ when 'grok_2_vision' then 32_768
+ else 131_072
+ end
+ end
+
+ def max_tokens_for(_model_id)
+ 4_096
+ end
+
+ def supports_vision?(model_id)
+ case model_family(model_id)
+ when 'grok_2_vision', 'grok_4' then true
+ else false
+ end
+ end
+
+ def supports_functions?(model_id)
+ model_family(model_id) != 'grok_2_image'
+ end
+
+ def supports_structured_output?(model_id)
+ model_family(model_id) != 'grok_2_image'
+ end
+
+ def supports_json_mode?(model_id)
+ supports_structured_output?(model_id)
+ end
+
+ # Pricing from API data (per million tokens)
+ # rubocop:disable Naming/VariableNumber
+ PRICES = {
+ grok_2: { input: 2.0, output: 10.0 },
+ grok_2_vision: { input: 2.0, output: 10.0 },
+ grok_3: { input: 3.0, output: 15.0, cached_input: 0.75 },
+ grok_3_fast: { input: 5.0, output: 25.0, cached_input: 1.25 },
+ grok_3_mini: { input: 0.3, output: 0.5, cached_input: 0.075 },
+ grok_3_mini_fast: { input: 0.6, output: 4.0, cached_input: 0.15 },
+ grok_4: { input: 3.0, output: 15.0, cached_input: 0.75 }
+ }.freeze
+ # rubocop:enable Naming/VariableNumber
+
+ def model_family(model_id)
+ MODEL_PATTERNS.each do |family, pattern|
+ return family.to_s if model_id.match?(pattern)
+ end
+ 'other'
+ end
+
+ def input_price_for(model_id)
+ family = model_family(model_id).to_sym
+ prices = PRICES.fetch(family, { input: default_input_price })
+ prices[:input] || default_input_price
+ end
+
+ def cached_input_price_for(model_id)
+ family = model_family(model_id).to_sym
+ prices = PRICES.fetch(family, {})
+ prices[:cached_input]
+ end
+
+ def output_price_for(model_id)
+ family = model_family(model_id).to_sym
+ prices = PRICES.fetch(family, { output: default_output_price })
+ prices[:output] || default_output_price
+ end
+
+ def model_type(model_id)
+ return 'image' if model_family(model_id) == 'grok_2_image'
+
+ 'chat'
+ end
+
+ def default_input_price
+ 2.0
+ end
+
+ def default_output_price
+ 10.0
+ end
+
+ def format_display_name(model_id)
+ model_id.then { |id| humanize(id) }
+ .then { |name| apply_special_formatting(name) }
+ end
+
+ def humanize(id)
+ id.tr('-', ' ')
+ .split
+ .map(&:capitalize)
+ .join(' ')
+ end
+
+ def apply_special_formatting(name)
+ name
+ .gsub(/^Grok /, 'Grok-')
+ .gsub(/(\d{4}) (\d{2}) (\d{2})/, '\1-\2-\3')
+ end
+
+ def modalities_for(model_id)
+ modalities = {
+ input: ['text'],
+ output: []
+ }
+
+ modalities[:output] << 'text' if model_type(model_id) == 'chat'
+
+ # Vision support
+ modalities[:input] << 'image' if supports_vision?(model_id)
+
+ modalities
+ end
+
+ def capabilities_for(model_id)
+ capabilities = []
+
+ # Common capabilities
+ capabilities << 'streaming'
+ capabilities << 'function_calling' if supports_functions?(model_id)
+ capabilities << 'structured_output' if supports_structured_output?(model_id)
+
+ capabilities
+ end
+
+ def pricing_for(model_id)
+ standard_pricing = {
+ input_per_million: input_price_for(model_id),
+ output_per_million: output_price_for(model_id)
+ }
+
+ # Add cached pricing if available
+ cached_price = cached_input_price_for(model_id)
+ standard_pricing[:cached_input_per_million] = cached_price if cached_price
+
+ # Pricing structure
+ { text_tokens: { standard: standard_pricing } }
+ end
+ end
+ end
+ end
+end
From 94af072d9c42755a1bd06ddddea47def10b167e3 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:04:29 +1000
Subject: [PATCH 05/28] Update Provider#list_models to support an array of URLs
---
lib/ruby_llm/provider.rb | 11 +++++++++--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/lib/ruby_llm/provider.rb b/lib/ruby_llm/provider.rb
index 5222f91f1..2b61635d7 100644
--- a/lib/ruby_llm/provider.rb
+++ b/lib/ruby_llm/provider.rb
@@ -60,8 +60,15 @@ def complete(messages, tools:, temperature:, model:, params: {}, headers: {}, sc
end
def list_models
- response = @connection.get models_url
- parse_list_models_response response, slug, capabilities
+ if models_url.is_a?(Array)
+ models_url.map do |url|
+ response = @connection.get(url)
+ parse_list_models_response response, slug, capabilities
+ end.flatten
+ else
+ response = @connection.get models_url
+ parse_list_models_response response, slug, capabilities
+ end
end
def embed(text, model:, dimensions:)
From f3193a4cdb4a60d7c0498d8e80c37a7fd4fc0f5a Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:04:44 +1000
Subject: [PATCH 06/28] Add xAI Models implementation
---
lib/ruby_llm/providers/xai.rb | 1 +
lib/ruby_llm/providers/xai/models.rb | 51 ++++++++++++++++++++++++++++
2 files changed, 52 insertions(+)
create mode 100644 lib/ruby_llm/providers/xai/models.rb
diff --git a/lib/ruby_llm/providers/xai.rb b/lib/ruby_llm/providers/xai.rb
index 3c157117d..b77015c7a 100644
--- a/lib/ruby_llm/providers/xai.rb
+++ b/lib/ruby_llm/providers/xai.rb
@@ -6,6 +6,7 @@ module Providers
class XAI < OpenAI
include XAI::Capabilities
include XAI::Chat
+ include XAI::Models
def api_base
'https://api.x.ai/v1'
diff --git a/lib/ruby_llm/providers/xai/models.rb b/lib/ruby_llm/providers/xai/models.rb
new file mode 100644
index 000000000..6cddca6f1
--- /dev/null
+++ b/lib/ruby_llm/providers/xai/models.rb
@@ -0,0 +1,51 @@
+# frozen_string_literal: true
+
+module RubyLLM
+ module Providers
+ class XAI
+ # Model definitions for xAI API
+ # https://docs.x.ai/docs/api-reference#list-language-models
+ # https://docs.x.ai/docs/api-reference#list-image-generation-models
+ module Models
+ module_function
+
+ # NOTE: We pull models list from two endpoints here as these provide
+ # detailed, modality, capability and cost information for each
+ # model that we can leverage which the generic OpenAI compatible
+ # /models endpoint does not provide.
+ def models_url
+ %w[language-models image-generation-models]
+ end
+
+ def parse_list_models_response(response, slug, capabilities)
+ data = response.body
+ return [] if data.empty?
+
+ data['models']&.map do |model_data|
+ # puts model_data
+ model_id = model_data['id']
+
+ info = Model::Info.new(
+ id: model_id,
+ name: capabilities.format_display_name(model_id),
+ provider: slug,
+ family: capabilities.model_family(model_id),
+ modalities: {
+ input: model_data['input_modalities'] | capabilities.modalities_for(model_id)[:input],
+ output: model_data['output_modalities'] | capabilities.modalities_for(model_id)[:output]
+ },
+ context_window: capabilities.context_window_for(model_id),
+ capabilities: capabilities.capabilities_for(model_id),
+ pricing: capabilities.pricing_for(model_id),
+ metadata: {
+ aliases: model_data['aliases']
+ }
+ )
+ puts info.inspect
+ info
+ end || []
+ end
+ end
+ end
+ end
+end
From 856e125d486cb69195ec513061ec6c19ba81274c Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:06:25 +1000
Subject: [PATCH 07/28] Alphabetise Providers in Models.rake config
---
lib/tasks/models.rake | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/lib/tasks/models.rake b/lib/tasks/models.rake
index 4cadb7bd0..9a80a1992 100644
--- a/lib/tasks/models.rake
+++ b/lib/tasks/models.rake
@@ -39,14 +39,16 @@ end
def configure_from_env
RubyLLM.configure do |config|
- config.openai_api_key = ENV.fetch('OPENAI_API_KEY', nil)
+ # Providers
config.anthropic_api_key = ENV.fetch('ANTHROPIC_API_KEY', nil)
- config.gemini_api_key = ENV.fetch('GEMINI_API_KEY', nil)
config.deepseek_api_key = ENV.fetch('DEEPSEEK_API_KEY', nil)
- config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', nil)
- config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', nil)
+ config.gemini_api_key = ENV.fetch('GEMINI_API_KEY', nil)
config.mistral_api_key = ENV.fetch('MISTRAL_API_KEY', nil)
+ config.openai_api_key = ENV.fetch('OPENAI_API_KEY', nil)
+ config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', nil)
+ config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', nil)
configure_bedrock(config)
+ # Requests
config.request_timeout = 30
end
end
From 882fbc3a720bded9a7589a6ef647534cc7b2dca5 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:06:40 +1000
Subject: [PATCH 08/28] Add xAI API key to models.rake and console config
---
bin/console | 1 +
lib/tasks/models.rake | 1 +
2 files changed, 2 insertions(+)
diff --git a/bin/console b/bin/console
index 3d0cb9176..6f07ae4fc 100755
--- a/bin/console
+++ b/bin/console
@@ -23,6 +23,7 @@ RubyLLM.configure do |config|
config.openai_api_key = ENV.fetch('OPENAI_API_KEY', nil)
config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', nil)
config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', nil)
+ config.xai_api_key = ENV.fetch('XAI_API_KEY', nil)
end
IRB.start(__FILE__)
diff --git a/lib/tasks/models.rake b/lib/tasks/models.rake
index 9a80a1992..7433fd070 100644
--- a/lib/tasks/models.rake
+++ b/lib/tasks/models.rake
@@ -47,6 +47,7 @@ def configure_from_env
config.openai_api_key = ENV.fetch('OPENAI_API_KEY', nil)
config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', nil)
config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', nil)
+ config.xai_api_key = ENV.fetch('XAI_API_KEY', nil)
configure_bedrock(config)
# Requests
config.request_timeout = 30
From 2b01de293656e7584480d3a14e0e4f77bbf9a639 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:06:53 +1000
Subject: [PATCH 09/28] Add xAI support to models.rake
---
lib/tasks/models.rake | 21 +++++++++++++++++++++
1 file changed, 21 insertions(+)
diff --git a/lib/tasks/models.rake b/lib/tasks/models.rake
index 7433fd070..5896d4d99 100644
--- a/lib/tasks/models.rake
+++ b/lib/tasks/models.rake
@@ -419,6 +419,27 @@ def generate_aliases # rubocop:disable Metrics/PerceivedComplexity
}
end
+ models['xai'].each do |model|
+ # xAI aliases
+ m = RubyLLM.models.find(model)
+ next unless m.metadata&.dig(:aliases)
+
+ m.metadata[:aliases].each do |alias_name|
+ aliases[alias_name] = { 'xai' => m.id }
+ end
+
+ # OpenRouter aliases.
+ # NOTE: OpenRouter uses "x-ai" as the provider slug
+ openrouter_model = "x-ai/#{model}"
+ next unless models['openrouter'].include?(openrouter_model)
+
+ alias_key = model.gsub('-latest', '').gsub(/-\d{4}/, '-4')
+ aliases[alias_key] = {
+ 'xai' => alias_key,
+ 'openrouter' => openrouter_model
+ }
+ end
+
sorted_aliases = aliases.sort.to_h
File.write(RubyLLM::Aliases.aliases_file, JSON.pretty_generate(sorted_aliases))
From c9483aaa86d48e7e8d10f38a1e9fc645a3c0f3f5 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:08:12 +1000
Subject: [PATCH 10/28] Update models, aliases and docs
---
docs/_reference/available-models.md | 493 +-
lib/ruby_llm/aliases.json | 114 +-
lib/ruby_llm/models.json | 19402 +++++++++++++-------------
3 files changed, 10238 insertions(+), 9771 deletions(-)
diff --git a/docs/_reference/available-models.md b/docs/_reference/available-models.md
index 90e632e22..3c169c387 100644
--- a/docs/_reference/available-models.md
+++ b/docs/_reference/available-models.md
@@ -47,70 +47,58 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
## Last Updated
{: .d-inline-block }
-2025-08-23
+2025-08-26
{: .label .label-green }
## Models by Provider
-### Anthropic (21)
+### Anthropic (9)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| anthropic.claude-3-haiku-20240307-v1:0 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
-| claude-3-haiku@20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
-| anthropic.claude-3-5-haiku-20241022-v1:0 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
-| claude-3-5-haiku@20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-opus-4-20250514-v1:0 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-opus-4-20250514 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| claude-opus-4@20250514 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| anthropic.claude-opus-4-1-20250805-v1:0 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-opus-4-1-20250805 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| claude-opus-4-1@20250805 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-7-sonnet-20250219-v1:0 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
| claude-3-7-sonnet-20250219 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| claude-3-7-sonnet@20250219 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| anthropic.claude-sonnet-4-20250514-v1:0 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
| claude-sonnet-4-20250514 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| claude-sonnet-4@20250514 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
### Bedrock (31)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| anthropic.claude-v2 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-v2:1 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-instant-v1 | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
+| anthropic.claude-instant-v1 | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| us.anthropic.claude-opus-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-opus-4-1-20250805-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-sonnet-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
@@ -124,36 +112,38 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| deepseek-reasoner | deepseek | - | 64000 | In: $0.55, Out: $2.19, Cache: $0.14 |
-### Gemini (58)
+### Gemini (57)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
| embedding-001 | gemini | 2048 | 1 | - |
| embedding-gecko-001 | gemini | 1024 | 1 | - |
+| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-latest | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
+| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
-| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash-exp-image-generation | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-live-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-preview-image-generation | gemini | 32000 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
-| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-preview | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-lite-preview-02-05 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp-02-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-image-preview | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemini-live-2.5-flash-preview | gemini | 1048576 | 8192 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-exp-native-audio-thinking-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-native-audio-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
@@ -180,9 +170,6 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gemma-3n-e2b-it | gemini | 8192 | 2048 | In: $0.08, Out: $0.30 |
| gemma-3n-e4b-it | gemini | 8192 | 2048 | In: $0.08, Out: $0.30 |
| imagen-3.0-generate-002 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-001 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-preview-06-06 | gemini | 480 | 8192 | - |
-| imagen-4.0-ultra-generate-preview-06-06 | gemini | 480 | 8192 | - |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
| aqa | gemini | 7168 | 1024 | - |
| text-embedding-004 | gemini | 2048 | 1 | - |
@@ -192,8 +179,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| codestral-2411-rc5 | mistral | 32768 | 8192 | - |
| codestral-2412 | mistral | 32768 | 8192 | - |
+| codestral-2411-rc5 | mistral | 32768 | 8192 | - |
| codestral-2501 | mistral | 32768 | 8192 | - |
| codestral-2508 | mistral | 32768 | 8192 | - |
| codestral-embed | mistral | 32768 | 8192 | - |
@@ -223,8 +210,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| mistral-medium-2505 | mistral | 32768 | 8192 | - |
| mistral-medium-2508 | mistral | 32768 | 8192 | - |
| mistral-medium-latest | mistral | 32768 | 8192 | - |
-| mistral-moderation-2411 | mistral | 32768 | 8192 | - |
| mistral-moderation-latest | mistral | 32768 | 8192 | - |
+| mistral-moderation-2411 | mistral | 32768 | 8192 | - |
| mistral-ocr-2503 | mistral | 32768 | 8192 | - |
| mistral-ocr-2505 | mistral | 32768 | 8192 | - |
| mistral-ocr-latest | mistral | 32768 | 8192 | - |
@@ -286,19 +273,20 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
-| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-realtime-preview | openai | 32000 | 4096 | In: $5.00, Out: $20.00, Cache: $2.50 |
-| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-transcribe | openai | 16000 | 2000 | In: $2.50, Out: $10.00 |
| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-4o-mini-realtime-preview | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
@@ -309,21 +297,20 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gpt-4o-audio-preview-2024-10-01 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2025-06-03 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-4o-realtime-preview-2024-10-01 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2025-06-03 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
-| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| gpt-5-chat-latest | openai | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
-| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| omni-moderation-2024-09-26 | openai | - | - | - |
| tts-1 | openai | - | - | In: $15.00 |
| tts-1-1106 | openai | - | - | In: $15.00, Out: $15.00 |
-| tts-1-hd | openai | - | - | In: $30.00 |
+| tts-1-hd | openai | - | - | Out: $30.00 |
| tts-1-hd-1106 | openai | - | - | In: $30.00, Out: $30.00 |
| whisper-1 | openai | - | - | In: $0.01 |
| babbage-002 | openai | - | 16384 | In: $0.40, Out: $0.40 |
@@ -339,14 +326,14 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| o1-preview-2024-09-12 | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-mini | openai | 128000 | 65536 | In: $1.10, Out: $4.40, Cache: $0.55 |
| o1-mini-2024-09-12 | openai | 128000 | 65536 | In: $1.10, Out: $4.40, Cache: $0.55 |
-| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
| o3-deep-research | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
| o3-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
-| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
| o3-pro | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
| o3-pro-2025-06-10 | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
| o4-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
@@ -361,7 +348,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| text-moderation-stable | openai | - | 32768 | - |
-### OpenRouter (316)
+### OpenRouter (322)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
@@ -411,8 +398,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| deepseek/deepseek-prover-v2 | openrouter | 163840 | - | In: $0.50, Out: $2.18 |
| deepseek/deepseek-chat | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-chat-v3-0324 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
-| deepseek/deepseek-chat-v3-0324:free | openrouter | 163840 | 163840 | - |
-| deepseek/deepseek-v3-base | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
+| deepseek/deepseek-chat-v3-0324:free | openrouter | 163840 | - | - |
| deepseek/deepseek-chat-v3.1 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-v3.1-base | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-r1-0528-qwen3-8b | openrouter | 32000 | - | In: $0.01, Out: $0.02 |
@@ -421,7 +407,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| deepseek/deepseek-r1:free | openrouter | 163840 | - | - |
| deepseek/deepseek-r1-0528 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-r1-0528:free | openrouter | 163840 | - | - |
-| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.13 |
+| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.10 |
| deepseek/deepseek-r1-distill-llama-70b:free | openrouter | 8192 | 4096 | - |
| deepseek/deepseek-r1-distill-llama-8b | openrouter | 32000 | 32000 | In: $0.04, Out: $0.04 |
| deepseek/deepseek-r1-distill-qwen-1.5b | openrouter | 131072 | 32768 | In: $0.18, Out: $0.18 |
@@ -442,6 +428,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| google/gemini-2.0-flash-exp:free | openrouter | 1048576 | 8192 | - |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
+| google/gemini-2.5-flash-image-preview:free | openrouter | 32768 | 8192 | - |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -469,8 +457,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| liquid/lfm-7b | openrouter | 32768 | - | In: $0.01, Out: $0.01 |
| meta-llama/llama-guard-3-8b | openrouter | 131072 | - | In: $0.02, Out: $0.06 |
| anthracite-org/magnum-v2-72b | openrouter | 32768 | - | In: $3.00, Out: $3.00 |
-| anthracite-org/magnum-v4-72b | openrouter | 16384 | 1024 | In: $2.00, Out: $5.00 |
-| mancer/weaver | openrouter | 8000 | 1000 | In: $1.12, Out: $1.12 |
+| anthracite-org/magnum-v4-72b | openrouter | 16384 | 2048 | In: $2.00, Out: $5.00 |
+| mancer/weaver | openrouter | 8000 | 2000 | In: $1.12, Out: $1.12 |
| meta-llama/llama-3-70b-instruct | openrouter | 8192 | 16384 | In: $0.30, Out: $0.40 |
| meta-llama/llama-3-8b-instruct | openrouter | 8192 | 16384 | In: $0.03, Out: $0.06 |
| meta-llama/llama-3.1-405b | openrouter | 32768 | - | In: $2.00, Out: $2.00 |
@@ -486,8 +474,11 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| meta-llama/llama-3.2-90b-vision-instruct | openrouter | 131072 | 2048 | In: $1.20, Out: $1.20 |
| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 16384 | In: $0.04, Out: $0.12 |
| meta-llama/llama-3.3-70b-instruct:free | openrouter | 65536 | - | - |
+| meta-llama/llama-3.3-8b-instruct:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-maverick | openrouter | 1048576 | 16384 | In: $0.15, Out: $0.60 |
+| meta-llama/llama-4-maverick:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-scout | openrouter | 1048576 | 1048576 | In: $0.08, Out: $0.30 |
+| meta-llama/llama-4-scout:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-guard-4-12b | openrouter | 163840 | - | In: $0.18, Out: $0.18 |
| meta-llama/llama-guard-2-8b | openrouter | 8192 | - | In: $0.20, Out: $0.20 |
| microsoft/mai-ds-r1 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
@@ -522,7 +513,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| mistralai/mistral-7b-instruct-v0.1 | openrouter | 2824 | - | In: $0.11, Out: $0.19 |
| mistralai/mistral-7b-instruct-v0.3 | openrouter | 32768 | 16384 | In: $0.03, Out: $0.05 |
| mistralai/mistral-medium-3 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
-| mistralai/mistral-medium-3.1 | openrouter | 262144 | - | In: $0.40, Out: $2.00 |
+| mistralai/mistral-medium-3.1 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
| mistralai/mistral-nemo | openrouter | 32000 | - | In: $0.01, Out: $0.05 |
| mistralai/mistral-nemo:free | openrouter | 131072 | 128000 | - |
| mistralai/mistral-small-24b-instruct-2501 | openrouter | 32768 | - | In: $0.02, Out: $0.08 |
@@ -556,6 +547,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| nousresearch/nous-hermes-2-mixtral-8x7b-dpo | openrouter | 32768 | 2048 | In: $0.60, Out: $0.60 |
| nousresearch/hermes-3-llama-3.1-405b | openrouter | 131072 | 16384 | In: $0.70, Out: $0.80 |
| nousresearch/hermes-3-llama-3.1-70b | openrouter | 131072 | - | In: $0.10, Out: $0.28 |
+| nousresearch/hermes-4-405b | openrouter | 131072 | - | In: $1.00, Out: $3.00 |
+| nousresearch/hermes-4-70b | openrouter | 131072 | - | In: $0.09, Out: $0.37 |
| nousresearch/hermes-2-pro-llama-3-8b | openrouter | 131072 | 131072 | In: $0.02, Out: $0.04 |
| openai/chatgpt-4o-latest | openrouter | 128000 | 16384 | In: $5.00, Out: $15.00 |
| openai/codex-mini | openrouter | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
@@ -606,9 +599,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| perplexity/sonar-reasoning | openrouter | 127000 | - | In: $1.00, Out: $5.00 |
| perplexity/sonar-reasoning-pro | openrouter | 128000 | - | In: $2.00, Out: $8.00 |
| pygmalionai/mythalion-13b | openrouter | 4096 | 4096 | In: $0.70, Out: $1.10 |
-| featherless/qwerky-72b:free | openrouter | 32768 | 4096 | - |
| qwen/qwen-2-72b-instruct | openrouter | 32768 | 4096 | In: $0.90, Out: $0.90 |
-| qwen/qwen-2.5-72b-instruct | openrouter | 32768 | - | In: $0.07, Out: $0.27 |
+| qwen/qwen-2.5-72b-instruct | openrouter | 32768 | - | In: $0.05, Out: $0.21 |
| qwen/qwen-2.5-72b-instruct:free | openrouter | 32768 | - | - |
| qwen/qwen-2.5-7b-instruct | openrouter | 65536 | - | In: $0.04, Out: $0.10 |
| qwen/qwen-2.5-coder-32b-instruct | openrouter | 32768 | - | In: $0.05, Out: $0.20 |
@@ -634,7 +626,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| qwen/qwen3-235b-a22b-thinking-2507 | openrouter | 262144 | - | In: $0.08, Out: $0.31 |
| qwen/qwen3-30b-a3b | openrouter | 40960 | - | In: $0.02, Out: $0.08 |
| qwen/qwen3-30b-a3b:free | openrouter | 40960 | - | - |
-| qwen/qwen3-30b-a3b-instruct-2507 | openrouter | 131072 | 32768 | In: $0.20, Out: $0.80 |
+| qwen/qwen3-30b-a3b-instruct-2507 | openrouter | 262144 | - | In: $0.10, Out: $0.30 |
| qwen/qwen3-32b | openrouter | 40960 | - | In: $0.02, Out: $0.07 |
| qwen/qwen3-4b:free | openrouter | 40960 | - | - |
| qwen/qwen3-8b | openrouter | 128000 | 20000 | In: $0.04, Out: $0.14 |
@@ -680,6 +672,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| x-ai/grok-3-mini | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-3-mini-beta | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| x-ai/grok-code-fast-1 | openrouter | 256000 | 10000 | In: $0.20, Out: $1.50, Cache: $0.02 |
| x-ai/grok-vision-beta | openrouter | 8192 | - | In: $5.00, Out: $15.00 |
@@ -694,92 +687,98 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| sonar-reasoning-pro | perplexity | 128000 | 8192 | In: $2.00, Out: $8.00 |
+### XAI (10)
+
+| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
+| :-- | :-- | --: | --: | :-- |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-fast | xai | 131072 | - | In: $5.00, Out: $25.00, Cache: $1.25 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-3-mini-fast | xai | 131072 | - | In: $0.60, Out: $4.00, Cache: $0.15 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+
+
## Models by Capability
-### Function Calling (341)
+### Function Calling (348)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| anthropic.claude-3-haiku-20240307-v1:0 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
-| claude-3-haiku@20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
-| anthropic.claude-3-5-haiku-20241022-v1:0 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
-| claude-3-5-haiku@20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-opus-4-20250514-v1:0 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-opus-4-20250514 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| claude-opus-4@20250514 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| anthropic.claude-opus-4-1-20250805-v1:0 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-opus-4-1-20250805 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| claude-opus-4-1@20250805 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-7-sonnet-20250219-v1:0 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
| claude-3-7-sonnet-20250219 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| claude-3-7-sonnet@20250219 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| anthropic.claude-sonnet-4-20250514-v1:0 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
| claude-sonnet-4-20250514 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| claude-sonnet-4@20250514 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
| anthropic.claude-v2 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
-| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-instant-v1 | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
+| anthropic.claude-instant-v1 | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| us.anthropic.claude-opus-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-opus-4-1-20250805-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-sonnet-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| deepseek-chat | deepseek | 128000 | 8000 | In: $0.27, Out: $1.10, Cache: $0.07 |
| deepseek-reasoner | deepseek | - | 64000 | In: $0.55, Out: $2.19, Cache: $0.14 |
-| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-pro-latest | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
+| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
-| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
-| gemini-1.5-pro-latest | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash-exp-image-generation | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-live-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
-| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-pro-exp | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp-02-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-image-preview | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemini-live-2.5-flash-preview | gemini | 1048576 | 8192 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.5-flash-exp-native-audio-thinking-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-native-audio-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.5-flash-exp-native-audio-thinking-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
@@ -791,11 +790,11 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-exp-1206 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
-| codestral-2411-rc5 | mistral | 32768 | 8192 | - |
| codestral-2412 | mistral | 32768 | 8192 | - |
-| codestral-2501 | mistral | 32768 | 8192 | - |
| codestral-2508 | mistral | 32768 | 8192 | - |
| codestral-latest | mistral | 32768 | 8192 | - |
+| codestral-2501 | mistral | 32768 | 8192 | - |
+| codestral-2411-rc5 | mistral | 32768 | 8192 | - |
| devstral-medium-2507 | mistral | 32768 | 8192 | - |
| devstral-medium-latest | mistral | 32768 | 8192 | - |
| devstral-small-2505 | mistral | 32768 | 8192 | - |
@@ -811,18 +810,18 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| ministral-3b-latest | mistral | 32768 | 8192 | - |
| ministral-8b-2410 | mistral | 32768 | 8192 | - |
| ministral-8b-latest | mistral | 32768 | 8192 | - |
-| mistral-large-2407 | mistral | 32768 | 8192 | - |
| mistral-large-2411 | mistral | 32768 | 8192 | - |
+| mistral-large-2407 | mistral | 32768 | 8192 | - |
| mistral-large-latest | mistral | 32768 | 8192 | - |
| mistral-large-pixtral-2411 | mistral | 32768 | 8192 | - |
-| mistral-medium | mistral | 32768 | 8192 | - |
| mistral-medium-2505 | mistral | 32768 | 8192 | - |
+| mistral-medium | mistral | 32768 | 8192 | - |
| mistral-medium-2508 | mistral | 32768 | 8192 | - |
| mistral-medium-latest | mistral | 32768 | 8192 | - |
| mistral-saba-2502 | mistral | 32768 | 8192 | - |
| mistral-saba-latest | mistral | 32768 | 8192 | - |
-| mistral-small | mistral | 32768 | 8192 | - |
| mistral-small-2409 | mistral | 32768 | 8192 | - |
+| mistral-small | mistral | 32768 | 8192 | - |
| mistral-small-2501 | mistral | 32768 | 8192 | - |
| mistral-small-2503 | mistral | 32768 | 8192 | - |
| mistral-small-2506 | mistral | 32768 | 8192 | - |
@@ -836,23 +835,23 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| open-mixtral-8x22b | mistral | 32768 | 8192 | - |
| open-mixtral-8x22b-2404 | mistral | 32768 | 8192 | - |
| open-mixtral-8x7b | mistral | 32768 | 8192 | - |
-| pixtral-12b | mistral | 32768 | 8192 | - |
| pixtral-12b-2409 | mistral | 32768 | 8192 | - |
+| pixtral-12b | mistral | 32768 | 8192 | - |
| pixtral-12b-latest | mistral | 32768 | 8192 | - |
-| pixtral-large-2411 | mistral | 32768 | 8192 | - |
| pixtral-large-latest | mistral | 32768 | 8192 | - |
-| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
+| pixtral-large-2411 | mistral | 32768 | 8192 | - |
| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
+| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
@@ -860,17 +859,18 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-realtime-preview | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-realtime-preview | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
-| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| codex-mini-latest | openai | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
-| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
| computer-use-preview-2025-03-11 | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
| gpt-oss-120b | openai | 131072 | 131072 | - |
| gpt-oss-20b | openai | 131072 | 131072 | - |
| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
@@ -912,11 +912,11 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| cohere/command-r-plus-08-2024 | openrouter | 128000 | 4000 | In: $2.50, Out: $10.00 |
| deepseek/deepseek-chat | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-chat-v3-0324 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
-| deepseek/deepseek-chat-v3-0324:free | openrouter | 163840 | 163840 | - |
+| deepseek/deepseek-chat-v3-0324:free | openrouter | 163840 | - | - |
| deepseek/deepseek-chat-v3.1 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-r1 | openrouter | 163840 | 163840 | In: $0.40, Out: $2.00 |
| deepseek/deepseek-r1-0528 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
-| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.13 |
+| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.10 |
| google/gemini-flash-1.5 | openrouter | 1000000 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| google/gemini-flash-1.5-8b | openrouter | 1000000 | 8192 | In: $0.04, Out: $0.15, Cache: $0.01 |
| google/gemini-pro-1.5 | openrouter | 2000000 | 8192 | In: $1.25, Out: $5.00 |
@@ -940,8 +940,11 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| meta-llama/llama-3.2-3b-instruct | openrouter | 20000 | 20000 | In: $0.00, Out: $0.01 |
| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 16384 | In: $0.04, Out: $0.12 |
| meta-llama/llama-3.3-70b-instruct:free | openrouter | 65536 | - | - |
+| meta-llama/llama-3.3-8b-instruct:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-maverick | openrouter | 1048576 | 16384 | In: $0.15, Out: $0.60 |
+| meta-llama/llama-4-maverick:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-scout | openrouter | 1048576 | 1048576 | In: $0.08, Out: $0.30 |
+| meta-llama/llama-4-scout:free | openrouter | 128000 | 4028 | - |
| microsoft/phi-3-medium-128k-instruct | openrouter | 128000 | - | In: $1.00, Out: $1.00 |
| microsoft/phi-3-mini-128k-instruct | openrouter | 128000 | - | In: $0.10, Out: $0.10 |
| microsoft/phi-3.5-mini-128k-instruct | openrouter | 128000 | - | In: $0.10, Out: $0.10 |
@@ -966,7 +969,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| mistralai/mistral-7b-instruct-v0.1 | openrouter | 2824 | - | In: $0.11, Out: $0.19 |
| mistralai/mistral-7b-instruct-v0.3 | openrouter | 32768 | 16384 | In: $0.03, Out: $0.05 |
| mistralai/mistral-medium-3 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
-| mistralai/mistral-medium-3.1 | openrouter | 262144 | - | In: $0.40, Out: $2.00 |
+| mistralai/mistral-medium-3.1 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
| mistralai/mistral-nemo | openrouter | 32000 | - | In: $0.01, Out: $0.05 |
| mistralai/mistral-small-24b-instruct-2501 | openrouter | 32768 | - | In: $0.02, Out: $0.08 |
| mistralai/mistral-small-3.1-24b-instruct | openrouter | 131072 | 96000 | In: $0.02, Out: $0.08 |
@@ -982,6 +985,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| moonshotai/kimi-k2:free | openrouter | 32768 | - | - |
| nvidia/llama-3.1-nemotron-70b-instruct | openrouter | 131072 | 16384 | In: $0.12, Out: $0.30 |
| nousresearch/hermes-3-llama-3.1-70b | openrouter | 131072 | - | In: $0.10, Out: $0.28 |
+| nousresearch/hermes-4-70b | openrouter | 131072 | - | In: $0.09, Out: $0.37 |
| openai/codex-mini | openrouter | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
| openai/gpt-3.5-turbo | openrouter | 16385 | 4096 | In: $0.50, Out: $1.50 |
| openai/gpt-3.5-turbo-0613 | openrouter | 4095 | 4096 | In: $1.00, Out: $2.00 |
@@ -1014,7 +1018,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| openai/o3-pro | openrouter | 200000 | 100000 | In: $20.00, Out: $80.00 |
| openai/o4-mini | openrouter | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
| openai/o4-mini-high | openrouter | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
-| qwen/qwen-2.5-72b-instruct | openrouter | 32768 | - | In: $0.07, Out: $0.27 |
+| qwen/qwen-2.5-72b-instruct | openrouter | 32768 | - | In: $0.05, Out: $0.21 |
| qwen/qwen-max | openrouter | 32768 | 8192 | In: $1.60, Out: $6.40, Cache: $0.64 |
| qwen/qwen-plus | openrouter | 131072 | 8192 | In: $0.40, Out: $1.20, Cache: $0.16 |
| qwen/qwen-turbo | openrouter | 1000000 | 8192 | In: $0.05, Out: $0.20, Cache: $0.02 |
@@ -1024,6 +1028,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | - | In: $0.08, Out: $0.31 |
| qwen/qwen3-235b-a22b-thinking-2507 | openrouter | 262144 | - | In: $0.08, Out: $0.31 |
| qwen/qwen3-30b-a3b | openrouter | 40960 | - | In: $0.02, Out: $0.08 |
+| qwen/qwen3-30b-a3b-instruct-2507 | openrouter | 262144 | - | In: $0.10, Out: $0.30 |
| qwen/qwen3-32b | openrouter | 40960 | - | In: $0.02, Out: $0.07 |
| qwen/qwen3-4b:free | openrouter | 40960 | - | - |
| qwen/qwen3-coder | openrouter | 262144 | - | In: $0.20, Out: $0.80 |
@@ -1041,21 +1046,32 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| x-ai/grok-3-mini | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-3-mini-beta | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| x-ai/grok-code-fast-1 | openrouter | 256000 | 10000 | In: $0.20, Out: $1.50, Cache: $0.02 |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-fast | xai | 131072 | - | In: $5.00, Out: $25.00, Cache: $1.25 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-3-mini-fast | xai | 131072 | - | In: $0.60, Out: $4.00, Cache: $0.15 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
-### Structured Output (313)
+### Structured Output (331)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
| deepseek-chat | deepseek | 128000 | 8000 | In: $0.27, Out: $1.10, Cache: $0.07 |
| deepseek-reasoner | deepseek | - | 64000 | In: $0.55, Out: $2.19, Cache: $0.14 |
-| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
@@ -1063,6 +1079,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash-exp-image-generation | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-live-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-preview-image-generation | gemini | 32000 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
@@ -1071,10 +1088,11 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gemini-2.0-pro-exp-02-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-image-preview | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemini-live-2.5-flash-preview | gemini | 1048576 | 8192 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
-| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-preview-06-17 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
@@ -1116,12 +1134,12 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| mistral-medium-latest | mistral | 32768 | 8192 | - |
| mistral-saba-2502 | mistral | 32768 | 8192 | - |
| mistral-saba-latest | mistral | 32768 | 8192 | - |
+| mistral-small-latest | mistral | 32768 | 8192 | - |
| mistral-small | mistral | 32768 | 8192 | - |
| mistral-small-2409 | mistral | 32768 | 8192 | - |
| mistral-small-2501 | mistral | 32768 | 8192 | - |
| mistral-small-2503 | mistral | 32768 | 8192 | - |
| mistral-small-2506 | mistral | 32768 | 8192 | - |
-| mistral-small-latest | mistral | 32768 | 8192 | - |
| mistral-tiny | mistral | 32768 | 8192 | - |
| mistral-tiny-2407 | mistral | 32768 | 8192 | - |
| mistral-tiny-latest | mistral | 32768 | 8192 | - |
@@ -1144,14 +1162,14 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-4o-mini-search-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-4o-mini-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
@@ -1163,8 +1181,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| codex-mini-latest | openai | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
| gpt-oss-120b | openai | 131072 | 131072 | - |
| gpt-oss-20b | openai | 131072 | 131072 | - |
-| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-preview | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-preview-2024-09-12 | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
@@ -1194,7 +1212,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| deepseek/deepseek-chat-v3.1 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-r1 | openrouter | 163840 | 163840 | In: $0.40, Out: $2.00 |
| deepseek/deepseek-r1-0528 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
-| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.13 |
+| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.10 |
| deepseek/deepseek-r1-distill-qwen-32b | openrouter | 131072 | 16384 | In: $0.08, Out: $0.15 |
| alpindale/goliath-120b | openrouter | 6144 | 512 | In: $4.00, Out: $5.50 |
| google/gemini-flash-1.5 | openrouter | 1000000 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
@@ -1204,6 +1222,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| google/gemini-2.0-flash-exp:free | openrouter | 1048576 | 8192 | - |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
+| google/gemini-2.5-flash-image-preview:free | openrouter | 32768 | 8192 | - |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -1234,8 +1254,11 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| meta-llama/llama-3.2-3b-instruct | openrouter | 20000 | 20000 | In: $0.00, Out: $0.01 |
| meta-llama/llama-3.2-90b-vision-instruct | openrouter | 131072 | 2048 | In: $1.20, Out: $1.20 |
| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 16384 | In: $0.04, Out: $0.12 |
+| meta-llama/llama-3.3-8b-instruct:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-maverick | openrouter | 1048576 | 16384 | In: $0.15, Out: $0.60 |
+| meta-llama/llama-4-maverick:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-scout | openrouter | 1048576 | 1048576 | In: $0.08, Out: $0.30 |
+| meta-llama/llama-4-scout:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-guard-4-12b | openrouter | 163840 | - | In: $0.18, Out: $0.18 |
| microsoft/phi-4 | openrouter | 16384 | - | In: $0.06, Out: $0.14 |
| microsoft/phi-4-multimodal-instruct | openrouter | 131072 | - | In: $0.05, Out: $0.10 |
@@ -1259,7 +1282,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| mistralai/mistral-7b-instruct:free | openrouter | 32768 | 16384 | - |
| mistralai/mistral-7b-instruct-v0.3 | openrouter | 32768 | 16384 | In: $0.03, Out: $0.05 |
| mistralai/mistral-medium-3 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
-| mistralai/mistral-medium-3.1 | openrouter | 262144 | - | In: $0.40, Out: $2.00 |
+| mistralai/mistral-medium-3.1 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
| mistralai/mistral-nemo | openrouter | 32000 | - | In: $0.01, Out: $0.05 |
| mistralai/mistral-small-24b-instruct-2501 | openrouter | 32768 | - | In: $0.02, Out: $0.08 |
| mistralai/mistral-small-3.1-24b-instruct | openrouter | 131072 | 96000 | In: $0.02, Out: $0.08 |
@@ -1317,7 +1340,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| openai/o3-pro | openrouter | 200000 | 100000 | In: $20.00, Out: $80.00 |
| openai/o4-mini | openrouter | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
| openai/o4-mini-high | openrouter | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
-| qwen/qwen-2.5-72b-instruct | openrouter | 32768 | - | In: $0.07, Out: $0.27 |
+| qwen/qwen-2.5-72b-instruct | openrouter | 32768 | - | In: $0.05, Out: $0.21 |
| qwen/qwen-2.5-7b-instruct | openrouter | 65536 | - | In: $0.04, Out: $0.10 |
| qwen/qwen-2.5-coder-32b-instruct | openrouter | 32768 | - | In: $0.05, Out: $0.20 |
| qwen/qwq-32b | openrouter | 131072 | - | In: $0.08, Out: $0.15 |
@@ -1337,7 +1360,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | - | In: $0.08, Out: $0.31 |
| qwen/qwen3-235b-a22b-thinking-2507 | openrouter | 262144 | - | In: $0.08, Out: $0.31 |
| qwen/qwen3-30b-a3b | openrouter | 40960 | - | In: $0.02, Out: $0.08 |
-| qwen/qwen3-30b-a3b-instruct-2507 | openrouter | 131072 | 32768 | In: $0.20, Out: $0.80 |
+| qwen/qwen3-30b-a3b-instruct-2507 | openrouter | 262144 | - | In: $0.10, Out: $0.30 |
| qwen/qwen3-32b | openrouter | 40960 | - | In: $0.02, Out: $0.07 |
| qwen/qwen3-4b:free | openrouter | 40960 | - | - |
| qwen/qwen3-coder | openrouter | 262144 | - | In: $0.20, Out: $0.80 |
@@ -1359,41 +1382,52 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| x-ai/grok-3-mini | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-3-mini-beta | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| x-ai/grok-code-fast-1 | openrouter | 256000 | 10000 | In: $0.20, Out: $1.50, Cache: $0.02 |
| x-ai/grok-vision-beta | openrouter | 8192 | - | In: $5.00, Out: $15.00 |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-fast | xai | 131072 | - | In: $5.00, Out: $25.00, Cache: $1.25 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-3-mini-fast | xai | 131072 | - | In: $0.60, Out: $4.00, Cache: $0.15 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
-### Streaming (459)
+### Streaming (473)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
-| anthropic.claude-v2 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:1 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
-| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-instant-v1 | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
@@ -1402,10 +1436,12 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| us.anthropic.claude-sonnet-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| embedding-001 | gemini | 2048 | 1 | - |
| embedding-gecko-001 | gemini | 1024 | 1 | - |
+| gemini-2.0-flash-exp-image-generation | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-lite-preview | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-lite-preview-02-05 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp-02-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-2.5-flash-image-preview | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
@@ -1424,9 +1460,6 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gemma-3n-e2b-it | gemini | 8192 | 2048 | In: $0.08, Out: $0.30 |
| gemma-3n-e4b-it | gemini | 8192 | 2048 | In: $0.08, Out: $0.30 |
| imagen-3.0-generate-002 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-001 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-preview-06-06 | gemini | 480 | 8192 | - |
-| imagen-4.0-ultra-generate-preview-06-06 | gemini | 480 | 8192 | - |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
| aqa | gemini | 7168 | 1024 | - |
| text-embedding-004 | gemini | 2048 | 1 | - |
@@ -1498,7 +1531,6 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gpt-4o-audio-preview-2024-10-01 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2025-06-03 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-4o-realtime-preview-2024-10-01 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2025-06-03 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
@@ -1550,8 +1582,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| deepseek/deepseek-prover-v2 | openrouter | 163840 | - | In: $0.50, Out: $2.18 |
| deepseek/deepseek-chat | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-chat-v3-0324 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
-| deepseek/deepseek-chat-v3-0324:free | openrouter | 163840 | 163840 | - |
-| deepseek/deepseek-v3-base | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
+| deepseek/deepseek-chat-v3-0324:free | openrouter | 163840 | - | - |
| deepseek/deepseek-chat-v3.1 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-v3.1-base | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-r1-0528-qwen3-8b | openrouter | 32000 | - | In: $0.01, Out: $0.02 |
@@ -1560,7 +1591,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| deepseek/deepseek-r1:free | openrouter | 163840 | - | - |
| deepseek/deepseek-r1-0528 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
| deepseek/deepseek-r1-0528:free | openrouter | 163840 | - | - |
-| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.13 |
+| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.10 |
| deepseek/deepseek-r1-distill-llama-70b:free | openrouter | 8192 | 4096 | - |
| deepseek/deepseek-r1-distill-llama-8b | openrouter | 32000 | 32000 | In: $0.04, Out: $0.04 |
| deepseek/deepseek-r1-distill-qwen-1.5b | openrouter | 131072 | 32768 | In: $0.18, Out: $0.18 |
@@ -1581,6 +1612,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| google/gemini-2.0-flash-exp:free | openrouter | 1048576 | 8192 | - |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
+| google/gemini-2.5-flash-image-preview:free | openrouter | 32768 | 8192 | - |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -1608,8 +1641,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| liquid/lfm-7b | openrouter | 32768 | - | In: $0.01, Out: $0.01 |
| meta-llama/llama-guard-3-8b | openrouter | 131072 | - | In: $0.02, Out: $0.06 |
| anthracite-org/magnum-v2-72b | openrouter | 32768 | - | In: $3.00, Out: $3.00 |
-| anthracite-org/magnum-v4-72b | openrouter | 16384 | 1024 | In: $2.00, Out: $5.00 |
-| mancer/weaver | openrouter | 8000 | 1000 | In: $1.12, Out: $1.12 |
+| anthracite-org/magnum-v4-72b | openrouter | 16384 | 2048 | In: $2.00, Out: $5.00 |
+| mancer/weaver | openrouter | 8000 | 2000 | In: $1.12, Out: $1.12 |
| meta-llama/llama-3-70b-instruct | openrouter | 8192 | 16384 | In: $0.30, Out: $0.40 |
| meta-llama/llama-3-8b-instruct | openrouter | 8192 | 16384 | In: $0.03, Out: $0.06 |
| meta-llama/llama-3.1-405b | openrouter | 32768 | - | In: $2.00, Out: $2.00 |
@@ -1625,8 +1658,11 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| meta-llama/llama-3.2-90b-vision-instruct | openrouter | 131072 | 2048 | In: $1.20, Out: $1.20 |
| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 16384 | In: $0.04, Out: $0.12 |
| meta-llama/llama-3.3-70b-instruct:free | openrouter | 65536 | - | - |
+| meta-llama/llama-3.3-8b-instruct:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-maverick | openrouter | 1048576 | 16384 | In: $0.15, Out: $0.60 |
+| meta-llama/llama-4-maverick:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-scout | openrouter | 1048576 | 1048576 | In: $0.08, Out: $0.30 |
+| meta-llama/llama-4-scout:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-guard-4-12b | openrouter | 163840 | - | In: $0.18, Out: $0.18 |
| meta-llama/llama-guard-2-8b | openrouter | 8192 | - | In: $0.20, Out: $0.20 |
| microsoft/mai-ds-r1 | openrouter | 163840 | - | In: $0.20, Out: $0.80 |
@@ -1661,7 +1697,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| mistralai/mistral-7b-instruct-v0.1 | openrouter | 2824 | - | In: $0.11, Out: $0.19 |
| mistralai/mistral-7b-instruct-v0.3 | openrouter | 32768 | 16384 | In: $0.03, Out: $0.05 |
| mistralai/mistral-medium-3 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
-| mistralai/mistral-medium-3.1 | openrouter | 262144 | - | In: $0.40, Out: $2.00 |
+| mistralai/mistral-medium-3.1 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
| mistralai/mistral-nemo | openrouter | 32000 | - | In: $0.01, Out: $0.05 |
| mistralai/mistral-nemo:free | openrouter | 131072 | 128000 | - |
| mistralai/mistral-small-24b-instruct-2501 | openrouter | 32768 | - | In: $0.02, Out: $0.08 |
@@ -1695,6 +1731,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| nousresearch/nous-hermes-2-mixtral-8x7b-dpo | openrouter | 32768 | 2048 | In: $0.60, Out: $0.60 |
| nousresearch/hermes-3-llama-3.1-405b | openrouter | 131072 | 16384 | In: $0.70, Out: $0.80 |
| nousresearch/hermes-3-llama-3.1-70b | openrouter | 131072 | - | In: $0.10, Out: $0.28 |
+| nousresearch/hermes-4-405b | openrouter | 131072 | - | In: $1.00, Out: $3.00 |
+| nousresearch/hermes-4-70b | openrouter | 131072 | - | In: $0.09, Out: $0.37 |
| nousresearch/hermes-2-pro-llama-3-8b | openrouter | 131072 | 131072 | In: $0.02, Out: $0.04 |
| openai/chatgpt-4o-latest | openrouter | 128000 | 16384 | In: $5.00, Out: $15.00 |
| openai/codex-mini | openrouter | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
@@ -1745,9 +1783,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| perplexity/sonar-reasoning | openrouter | 127000 | - | In: $1.00, Out: $5.00 |
| perplexity/sonar-reasoning-pro | openrouter | 128000 | - | In: $2.00, Out: $8.00 |
| pygmalionai/mythalion-13b | openrouter | 4096 | 4096 | In: $0.70, Out: $1.10 |
-| featherless/qwerky-72b:free | openrouter | 32768 | 4096 | - |
| qwen/qwen-2-72b-instruct | openrouter | 32768 | 4096 | In: $0.90, Out: $0.90 |
-| qwen/qwen-2.5-72b-instruct | openrouter | 32768 | - | In: $0.07, Out: $0.27 |
+| qwen/qwen-2.5-72b-instruct | openrouter | 32768 | - | In: $0.05, Out: $0.21 |
| qwen/qwen-2.5-72b-instruct:free | openrouter | 32768 | - | - |
| qwen/qwen-2.5-7b-instruct | openrouter | 65536 | - | In: $0.04, Out: $0.10 |
| qwen/qwen-2.5-coder-32b-instruct | openrouter | 32768 | - | In: $0.05, Out: $0.20 |
@@ -1773,7 +1810,7 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| qwen/qwen3-235b-a22b-thinking-2507 | openrouter | 262144 | - | In: $0.08, Out: $0.31 |
| qwen/qwen3-30b-a3b | openrouter | 40960 | - | In: $0.02, Out: $0.08 |
| qwen/qwen3-30b-a3b:free | openrouter | 40960 | - | - |
-| qwen/qwen3-30b-a3b-instruct-2507 | openrouter | 131072 | 32768 | In: $0.20, Out: $0.80 |
+| qwen/qwen3-30b-a3b-instruct-2507 | openrouter | 262144 | - | In: $0.10, Out: $0.30 |
| qwen/qwen3-32b | openrouter | 40960 | - | In: $0.02, Out: $0.07 |
| qwen/qwen3-4b:free | openrouter | 40960 | - | - |
| qwen/qwen3-8b | openrouter | 128000 | 20000 | In: $0.04, Out: $0.14 |
@@ -1819,15 +1856,26 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| x-ai/grok-3-mini | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-3-mini-beta | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| x-ai/grok-code-fast-1 | openrouter | 256000 | 10000 | In: $0.20, Out: $1.50, Cache: $0.02 |
| x-ai/grok-vision-beta | openrouter | 8192 | - | In: $5.00, Out: $15.00 |
| sonar | perplexity | 128000 | 4096 | In: $1.00, Out: $1.00 |
| sonar-deep-research | perplexity | 128000 | 4096 | In: $2.00, Out: $8.00 |
| sonar-pro | perplexity | 200000 | 8192 | In: $3.00, Out: $15.00 |
| sonar-reasoning | perplexity | 128000 | 4096 | In: $1.00, Out: $5.00 |
| sonar-reasoning-pro | perplexity | 128000 | 8192 | In: $2.00, Out: $8.00 |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-fast | xai | 131072 | - | In: $5.00, Out: $25.00, Cache: $1.25 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-3-mini-fast | xai | 131072 | - | In: $0.60, Out: $4.00, Cache: $0.15 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
-### Batch Processing (125)
+### Batch Processing (127)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
@@ -1837,20 +1885,22 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| embedding-001 | gemini | 2048 | 1 | - |
| embedding-gecko-001 | gemini | 1024 | 1 | - |
+| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
-| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash-exp-image-generation | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-preview | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-lite-preview-02-05 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-image-preview | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
-| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-preview-06-17 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro | gemini | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| gemini-embedding-001 | gemini | 2048 | 1 | - |
@@ -1874,8 +1924,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| magistral-small-2506 | mistral | 32768 | 8192 | - |
| magistral-small-2507 | mistral | 32768 | 8192 | - |
| magistral-small-latest | mistral | 32768 | 8192 | - |
-| ministral-3b-2410 | mistral | 32768 | 8192 | - |
| ministral-3b-latest | mistral | 32768 | 8192 | - |
+| ministral-3b-2410 | mistral | 32768 | 8192 | - |
| ministral-8b-2410 | mistral | 32768 | 8192 | - |
| ministral-8b-latest | mistral | 32768 | 8192 | - |
| mistral-large-2407 | mistral | 32768 | 8192 | - |
@@ -1888,11 +1938,11 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| mistral-medium-latest | mistral | 32768 | 8192 | - |
| mistral-saba-2502 | mistral | 32768 | 8192 | - |
| mistral-saba-latest | mistral | 32768 | 8192 | - |
+| mistral-small-2503 | mistral | 32768 | 8192 | - |
| mistral-small | mistral | 32768 | 8192 | - |
| mistral-small-2312 | mistral | 32768 | 8192 | - |
| mistral-small-2409 | mistral | 32768 | 8192 | - |
| mistral-small-2501 | mistral | 32768 | 8192 | - |
-| mistral-small-2503 | mistral | 32768 | 8192 | - |
| mistral-small-2506 | mistral | 32768 | 8192 | - |
| mistral-small-latest | mistral | 32768 | 8192 | - |
| mistral-tiny | mistral | 32768 | 8192 | - |
@@ -1911,16 +1961,16 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| pixtral-large-2411 | mistral | 32768 | 8192 | - |
| pixtral-large-latest | mistral | 32768 | 8192 | - |
| gpt-3.5-turbo | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
-| gpt-4 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
| gpt-4-0613 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
+| gpt-4 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
@@ -1929,8 +1979,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
@@ -1939,8 +1989,8 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
| gpt-oss-20b | openai | 131072 | 131072 | - |
| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
-| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
| o3-deep-research | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
@@ -1960,71 +2010,59 @@ See the [Contributing Guide](https://github.com/crmne/ruby_llm/blob/main/CONTRIB
## Models by Modality
-### Vision Models (236)
+### Vision Models (231)
Models that can process images:
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| anthropic.claude-3-haiku-20240307-v1:0 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
-| claude-3-haiku@20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
-| anthropic.claude-3-5-haiku-20241022-v1:0 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
-| claude-3-5-haiku@20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-opus-4-20250514-v1:0 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-opus-4-20250514 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| claude-opus-4@20250514 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| anthropic.claude-opus-4-1-20250805-v1:0 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-opus-4-1-20250805 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| claude-opus-4-1@20250805 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-7-sonnet-20250219-v1:0 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
| claude-3-7-sonnet-20250219 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| claude-3-7-sonnet@20250219 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| anthropic.claude-sonnet-4-20250514-v1:0 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
| claude-sonnet-4-20250514 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| claude-sonnet-4@20250514 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
+| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
-| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-instant-v1 | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
+| anthropic.claude-instant-v1 | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| us.anthropic.claude-opus-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-opus-4-1-20250805-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-sonnet-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
-| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
@@ -2032,6 +2070,7 @@ Models that can process images:
| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash-exp-image-generation | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-preview-image-generation | gemini | 32000 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
@@ -2041,11 +2080,12 @@ Models that can process images:
| gemini-2.0-pro-exp-02-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.5-flash-image-preview | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
-| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-preview-06-17 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro | gemini | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| gemini-2.5-pro-preview-06-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
@@ -2055,9 +2095,6 @@ Models that can process images:
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-exp-1206 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| imagen-3.0-generate-002 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-001 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-preview-06-06 | gemini | 480 | 8192 | - |
-| imagen-4.0-ultra-generate-preview-06-06 | gemini | 480 | 8192 | - |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
| mistral-large-pixtral-2411 | mistral | 32768 | 8192 | - |
| pixtral-12b | mistral | 32768 | 8192 | - |
@@ -2067,8 +2104,8 @@ Models that can process images:
| pixtral-large-latest | mistral | 32768 | 8192 | - |
| chatgpt-4o-latest | openai | 128000 | 16384 | In: $5.00, Out: $15.00 |
| gpt-image-1 | openai | - | - | In: $5.00, Out: $40.00, Cache: $1.25 |
-| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
+| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
@@ -2077,27 +2114,27 @@ Models that can process images:
| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| gpt-5-chat-latest | openai | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
-| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| omni-moderation-2024-09-26 | openai | - | - | - |
| codex-mini-latest | openai | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
-| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
| computer-use-preview-2025-03-11 | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
-| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
-| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
| o3-deep-research | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
@@ -2133,6 +2170,8 @@ Models that can process images:
| google/gemini-2.0-flash-exp:free | openrouter | 1048576 | 8192 | - |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
+| google/gemini-2.5-flash-image-preview:free | openrouter | 32768 | 8192 | - |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -2149,12 +2188,14 @@ Models that can process images:
| meta-llama/llama-3.2-11b-vision-instruct:free | openrouter | 131072 | 2048 | - |
| meta-llama/llama-3.2-90b-vision-instruct | openrouter | 131072 | 2048 | In: $1.20, Out: $1.20 |
| meta-llama/llama-4-maverick | openrouter | 1048576 | 16384 | In: $0.15, Out: $0.60 |
+| meta-llama/llama-4-maverick:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-scout | openrouter | 1048576 | 1048576 | In: $0.08, Out: $0.30 |
+| meta-llama/llama-4-scout:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-guard-4-12b | openrouter | 163840 | - | In: $0.18, Out: $0.18 |
| microsoft/phi-4-multimodal-instruct | openrouter | 131072 | - | In: $0.05, Out: $0.10 |
| minimax/minimax-01 | openrouter | 1000192 | 1000192 | In: $0.20, Out: $1.10 |
| mistralai/mistral-medium-3 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
-| mistralai/mistral-medium-3.1 | openrouter | 262144 | - | In: $0.40, Out: $2.00 |
+| mistralai/mistral-medium-3.1 | openrouter | 131072 | - | In: $0.40, Out: $2.00 |
| mistralai/mistral-small-3.1-24b-instruct | openrouter | 131072 | 96000 | In: $0.02, Out: $0.08 |
| mistralai/mistral-small-3.1-24b-instruct:free | openrouter | 128000 | - | - |
| mistralai/mistral-small-3.2-24b-instruct | openrouter | 128000 | - | In: $0.05, Out: $0.10 |
@@ -2202,6 +2243,10 @@ Models that can process images:
| x-ai/grok-2-vision-1212 | openrouter | 32768 | - | In: $2.00, Out: $10.00 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
| x-ai/grok-vision-beta | openrouter | 8192 | - | In: $5.00, Out: $15.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
### Audio Input Models (49)
@@ -2210,17 +2255,17 @@ Models that can process audio:
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
+| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-latest | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
+| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
@@ -2231,8 +2276,8 @@ Models that can process audio:
| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-live-2.5-flash-preview | gemini | 1048576 | 8192 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.5-flash-exp-native-audio-thinking-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-native-audio-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-exp-native-audio-thinking-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-pro | gemini | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -2240,13 +2285,13 @@ Models that can process audio:
| gpt-4o-realtime-preview | openai | 32000 | 4096 | In: $5.00, Out: $20.00, Cache: $2.50 |
| gpt-4o-transcribe | openai | 16000 | 2000 | In: $2.50, Out: $10.00 |
| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-realtime-preview | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-realtime-preview | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
| gpt-4o-mini-transcribe | openai | 16000 | 2000 | In: $1.25, Out: $5.00 |
| gpt-4o-audio-preview-2024-10-01 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2025-06-03 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| tts-1-1106 | openai | - | - | In: $15.00, Out: $15.00 |
| tts-1-hd-1106 | openai | - | - | In: $30.00, Out: $30.00 |
| whisper-1 | openai | - | - | In: $0.01 |
@@ -2261,7 +2306,7 @@ Models that can process audio:
| openai/gpt-4o-audio-preview | openrouter | 128000 | 16384 | In: $2.50, Out: $10.00 |
-### PDF Models (56)
+### PDF Models (55)
Models that can process PDF documents:
@@ -2270,41 +2315,43 @@ Models that can process PDF documents:
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
-| anthropic.claude-v2 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:1 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1 | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
-| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-instant-v1 | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
+| anthropic.claude-instant-v1 | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| us.anthropic.claude-opus-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-opus-4-1-20250805-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-sonnet-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
+| gemini-2.0-flash-exp-image-generation | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-lite-preview | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-lite-preview-02-05 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp-02-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-2.5-flash-image-preview | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
@@ -2316,16 +2363,13 @@ Models that can process PDF documents:
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-exp-1206 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| imagen-3.0-generate-002 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-001 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-preview-06-06 | gemini | 480 | 8192 | - |
-| imagen-4.0-ultra-generate-preview-06-06 | gemini | 480 | 8192 | - |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| omni-moderation-2024-09-26 | openai | - | - | - |
-### Embedding Models (86)
+### Embedding Models (87)
Models that generate embeddings:
@@ -2345,29 +2389,30 @@ Models that generate embeddings:
| dall-e-3 | openai | - | - | - |
| gpt-image-1 | openai | - | - | In: $5.00, Out: $40.00, Cache: $1.25 |
| gpt-3.5-turbo | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
-| gpt-4 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
| gpt-4-0613 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
+| gpt-4 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4-turbo-preview | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
-| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
-| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-audio-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-realtime-preview | openai | 32000 | 4096 | In: $5.00, Out: $20.00, Cache: $2.50 |
-| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-transcribe | openai | 16000 | 2000 | In: $2.50, Out: $10.00 |
| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-4o-mini-realtime-preview | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
| gpt-4o-mini-search-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
@@ -2377,40 +2422,40 @@ Models that generate embeddings:
| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| gpt-5-chat-latest | openai | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| tts-1 | openai | - | - | In: $15.00 |
-| tts-1-hd | openai | - | - | In: $30.00 |
+| tts-1-hd | openai | - | - | Out: $30.00 |
| whisper-1 | openai | - | - | In: $0.01 |
| babbage-002 | openai | - | 16384 | In: $0.40, Out: $0.40 |
| codex-mini-latest | openai | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
-| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
| computer-use-preview-2025-03-11 | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
| davinci-002 | openai | - | 16384 | In: $2.00, Out: $2.00 |
| gpt-oss-120b | openai | 131072 | 131072 | - |
| gpt-oss-20b | openai | 131072 | 131072 | - |
| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
-| o1-preview | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-preview-2024-09-12 | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
-| o1-mini | openai | 128000 | 65536 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o1-preview | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-mini-2024-09-12 | openai | 128000 | 65536 | In: $1.10, Out: $4.40, Cache: $0.55 |
-| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o1-mini | openai | 128000 | 65536 | In: $1.10, Out: $4.40, Cache: $0.55 |
| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
-| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
-| o3-deep-research | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
| o3-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3-deep-research | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
-| o3-pro | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
| o3-pro-2025-06-10 | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
-| o4-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o3-pro | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
| o4-mini-2025-04-16 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
-| o4-mini-deep-research | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o4-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
| o4-mini-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o4-mini-deep-research | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
| omni-moderation-latest | openai | - | - | - |
| text-embedding-3-large | openai | - | - | In: $0.13 |
| text-embedding-3-small | openai | - | - | In: $0.02 |
diff --git a/lib/ruby_llm/aliases.json b/lib/ruby_llm/aliases.json
index 970c0e1f5..145d51c07 100644
--- a/lib/ruby_llm/aliases.json
+++ b/lib/ruby_llm/aliases.json
@@ -1,22 +1,4 @@
{
- "anthropic.claude-3-5-haiku-20241022-v1:0": {
- "anthropic": "anthropic.claude-3-5-haiku-20241022-v1:0"
- },
- "anthropic.claude-3-7-sonnet-20250219-v1:0": {
- "anthropic": "anthropic.claude-3-7-sonnet-20250219-v1:0"
- },
- "anthropic.claude-3-haiku-20240307-v1:0": {
- "anthropic": "anthropic.claude-3-haiku-20240307-v1:0"
- },
- "anthropic.claude-opus-4-1-20250805-v1:0": {
- "anthropic": "anthropic.claude-opus-4-1-20250805-v1:0"
- },
- "anthropic.claude-opus-4-20250514-v1:0": {
- "anthropic": "anthropic.claude-opus-4-20250514-v1:0"
- },
- "anthropic.claude-sonnet-4-20250514-v1:0": {
- "anthropic": "anthropic.claude-sonnet-4-20250514-v1:0"
- },
"chatgpt-4o": {
"openai": "chatgpt-4o-latest",
"openrouter": "openai/chatgpt-4o-latest"
@@ -26,9 +8,6 @@
"openrouter": "anthropic/claude-3.5-haiku",
"bedrock": "anthropic.claude-3-5-haiku-20241022-v1:0"
},
- "claude-3-5-haiku@20241022": {
- "anthropic": "claude-3-5-haiku@20241022"
- },
"claude-3-5-sonnet": {
"anthropic": "claude-3-5-sonnet-20241022",
"openrouter": "anthropic/claude-3.5-sonnet",
@@ -39,17 +18,11 @@
"openrouter": "anthropic/claude-3.7-sonnet",
"bedrock": "us.anthropic.claude-3-7-sonnet-20250219-v1:0"
},
- "claude-3-7-sonnet@20250219": {
- "anthropic": "claude-3-7-sonnet@20250219"
- },
"claude-3-haiku": {
"anthropic": "claude-3-haiku-20240307",
"openrouter": "anthropic/claude-3-haiku",
"bedrock": "anthropic.claude-3-haiku-20240307-v1:0:200k"
},
- "claude-3-haiku@20240307": {
- "anthropic": "claude-3-haiku@20240307"
- },
"claude-3-opus": {
"anthropic": "claude-3-opus-20240229",
"openrouter": "anthropic/claude-3-opus",
@@ -68,20 +41,11 @@
"openrouter": "anthropic/claude-opus-4.1",
"bedrock": "us.anthropic.claude-opus-4-1-20250805-v1:0"
},
- "claude-opus-4-1@20250805": {
- "anthropic": "claude-opus-4-1@20250805"
- },
- "claude-opus-4@20250514": {
- "anthropic": "claude-opus-4@20250514"
- },
"claude-sonnet-4": {
"anthropic": "claude-sonnet-4-20250514",
"openrouter": "anthropic/claude-sonnet-4",
"bedrock": "us.anthropic.claude-sonnet-4-20250514-v1:0"
},
- "claude-sonnet-4@20250514": {
- "anthropic": "claude-sonnet-4@20250514"
- },
"deepseek-chat": {
"deepseek": "deepseek-chat",
"openrouter": "deepseek/deepseek-chat"
@@ -98,6 +62,10 @@
"gemini": "gemini-2.5-flash",
"openrouter": "google/gemini-2.5-flash"
},
+ "gemini-2.5-flash-image-preview": {
+ "gemini": "gemini-2.5-flash-image-preview",
+ "openrouter": "google/gemini-2.5-flash-image-preview"
+ },
"gemini-2.5-flash-lite": {
"gemini": "gemini-2.5-flash-lite",
"openrouter": "google/gemini-2.5-flash-lite"
@@ -226,6 +194,80 @@
"openai": "gpt-oss-20b",
"openrouter": "openai/gpt-oss-20b"
},
+ "grok-2": {
+ "xai": "grok-2-1212"
+ },
+ "grok-2-4": {
+ "xai": "grok-2-4",
+ "openrouter": "x-ai/grok-2-1212"
+ },
+ "grok-2-image": {
+ "xai": "grok-2-image-1212"
+ },
+ "grok-2-image-latest": {
+ "xai": "grok-2-image-1212"
+ },
+ "grok-2-latest": {
+ "xai": "grok-2-1212"
+ },
+ "grok-2-vision": {
+ "xai": "grok-2-vision-1212"
+ },
+ "grok-2-vision-4": {
+ "xai": "grok-2-vision-4",
+ "openrouter": "x-ai/grok-2-vision-1212"
+ },
+ "grok-2-vision-latest": {
+ "xai": "grok-2-vision-1212"
+ },
+ "grok-3": {
+ "xai": "grok-3",
+ "openrouter": "x-ai/grok-3"
+ },
+ "grok-3-beta": {
+ "xai": "grok-3"
+ },
+ "grok-3-fast-beta": {
+ "xai": "grok-3-fast"
+ },
+ "grok-3-fast-latest": {
+ "xai": "grok-3-fast"
+ },
+ "grok-3-latest": {
+ "xai": "grok-3"
+ },
+ "grok-3-mini": {
+ "xai": "grok-3-mini",
+ "openrouter": "x-ai/grok-3-mini"
+ },
+ "grok-3-mini-beta": {
+ "xai": "grok-3-mini"
+ },
+ "grok-3-mini-fast-beta": {
+ "xai": "grok-3-mini-fast"
+ },
+ "grok-3-mini-fast-latest": {
+ "xai": "grok-3-mini-fast"
+ },
+ "grok-3-mini-latest": {
+ "xai": "grok-3-mini"
+ },
+ "grok-4": {
+ "xai": "grok-4-0709"
+ },
+ "grok-4-latest": {
+ "xai": "grok-4-0709"
+ },
+ "grok-code-fast": {
+ "xai": "grok-code-fast-1"
+ },
+ "grok-code-fast-1": {
+ "xai": "grok-code-fast-1",
+ "openrouter": "x-ai/grok-code-fast-1"
+ },
+ "grok-code-fast-1-0825": {
+ "xai": "grok-code-fast-1"
+ },
"o1": {
"openai": "o1",
"openrouter": "openai/o1"
diff --git a/lib/ruby_llm/models.json b/lib/ruby_llm/models.json
index 514f9633d..6e88d3ae8 100644
--- a/lib/ruby_llm/models.json
+++ b/lib/ruby_llm/models.json
@@ -1,6 +1,6 @@
[
{
- "id": "anthropic.claude-3-5-haiku-20241022-v1:0",
+ "id": "claude-3-5-haiku-20241022",
"name": "Claude Haiku 3.5",
"provider": "anthropic",
"family": "claude-3-5-haiku",
@@ -32,7 +32,83 @@
"metadata": {}
},
{
- "id": "anthropic.claude-3-7-sonnet-20250219-v1:0",
+ "id": "claude-3-5-sonnet-20240620",
+ "name": "Claude Sonnet 3.5 (Old)",
+ "provider": "anthropic",
+ "family": "claude-3-5-sonnet",
+ "created_at": "2024-06-20 00:00:00 UTC",
+ "context_window": 200000,
+ "max_output_tokens": 8192,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image",
+ "pdf"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "batch"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
+ }
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "id": "claude-3-5-sonnet-20241022",
+ "name": "Claude Sonnet 3.5 (New)",
+ "provider": "anthropic",
+ "family": "claude-3-5-sonnet",
+ "created_at": "2024-10-22 00:00:00 UTC",
+ "context_window": 200000,
+ "max_output_tokens": 8192,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image",
+ "pdf"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "batch"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
+ }
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "id": "claude-3-7-sonnet-20250219",
"name": "Claude Sonnet 3.7",
"provider": "anthropic",
"family": "claude-3-7-sonnet",
@@ -64,7 +140,7 @@
"metadata": {}
},
{
- "id": "anthropic.claude-3-haiku-20240307-v1:0",
+ "id": "claude-3-haiku-20240307",
"name": "Claude Haiku 3",
"provider": "anthropic",
"family": "claude-3-haiku",
@@ -96,42 +172,48 @@
"metadata": {}
},
{
- "id": "anthropic.claude-opus-4-1-20250805-v1:0",
- "name": "Claude Opus 4.1",
+ "id": "claude-3-opus-20240229",
+ "name": "Claude Opus 3",
"provider": "anthropic",
- "family": "claude-opus-4-1",
- "created_at": null,
+ "family": "claude-3-opus",
+ "created_at": "2024-02-29 00:00:00 UTC",
"context_window": 200000,
- "max_output_tokens": 32000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "function_calling"
+ "streaming",
+ "function_calling",
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 15.0,
- "cached_input_per_million": 18.75,
"output_per_million": 75.0
+ },
+ "batch": {
+ "input_per_million": 7.5,
+ "output_per_million": 37.5
}
}
},
"metadata": {}
},
{
- "id": "anthropic.claude-opus-4-20250514-v1:0",
- "name": "Claude Opus 4",
+ "id": "claude-opus-4-1-20250805",
+ "name": "Claude Opus 4.1",
"provider": "anthropic",
- "family": "claude-opus-4",
+ "family": "claude-opus-4-1",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 32000,
@@ -160,13 +242,13 @@
"metadata": {}
},
{
- "id": "anthropic.claude-sonnet-4-20250514-v1:0",
- "name": "Claude Sonnet 4",
+ "id": "claude-opus-4-20250514",
+ "name": "Claude Opus 4",
"provider": "anthropic",
- "family": "claude-sonnet-4",
+ "family": "claude-opus-4",
"created_at": null,
"context_window": 200000,
- "max_output_tokens": 64000,
+ "max_output_tokens": 32000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -183,22 +265,22 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "cached_input_per_million": 3.75,
- "output_per_million": 15.0
+ "input_per_million": 15.0,
+ "cached_input_per_million": 18.75,
+ "output_per_million": 75.0
}
}
},
"metadata": {}
},
{
- "id": "claude-3-5-haiku-20241022",
- "name": "Claude Haiku 3.5",
+ "id": "claude-sonnet-4-20250514",
+ "name": "Claude Sonnet 4",
"provider": "anthropic",
- "family": "claude-3-5-haiku",
+ "family": "claude-sonnet-4",
"created_at": null,
"context_window": 200000,
- "max_output_tokens": 8192,
+ "max_output_tokens": 64000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -215,54 +297,71 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.8,
- "cached_input_per_million": 1.0,
- "output_per_million": 4.0
+ "input_per_million": 3.0,
+ "cached_input_per_million": 3.75,
+ "output_per_million": 15.0
}
}
},
"metadata": {}
},
{
- "id": "claude-3-5-haiku@20241022",
- "name": "Claude Haiku 3.5",
- "provider": "anthropic",
- "family": "claude-3-5-haiku",
+ "id": "anthropic.claude-3-5-haiku-20241022-v1:0",
+ "name": "Claude 3.5 Haiku",
+ "provider": "bedrock",
+ "family": "claude3_5_haiku",
"created_at": null,
"context_window": 200000,
- "max_output_tokens": 8192,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.8,
- "cached_input_per_million": 1.0,
"output_per_million": 4.0
+ },
+ "batch": {
+ "input_per_million": 0.4,
+ "output_per_million": 2.0
}
}
},
- "metadata": {}
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "ON_DEMAND"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
},
{
- "id": "claude-3-5-sonnet-20240620",
- "name": "Claude Sonnet 3.5 (Old)",
- "provider": "anthropic",
- "family": "claude-3-5-sonnet",
- "created_at": "2024-06-20 00:00:00 UTC",
+ "id": "anthropic.claude-3-5-sonnet-20240620-v1:0",
+ "name": "Claude 3.5 Sonnet",
+ "provider": "bedrock",
+ "family": "claude3_sonnet",
+ "created_at": null,
"context_window": 200000,
- "max_output_tokens": 8192,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -276,8 +375,7 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "batch"
+ "function_calling"
],
"pricing": {
"text_tokens": {
@@ -291,16 +389,30 @@
}
}
},
- "metadata": {}
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "ON_DEMAND",
+ "INFERENCE_PROFILE"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
},
{
- "id": "claude-3-5-sonnet-20241022",
- "name": "Claude Sonnet 3.5 (New)",
- "provider": "anthropic",
- "family": "claude-3-5-sonnet",
- "created_at": "2024-10-22 00:00:00 UTC",
+ "id": "anthropic.claude-3-5-sonnet-20240620-v1:0:18k",
+ "name": "Claude 3.5 Sonnet",
+ "provider": "bedrock",
+ "family": "claude3_sonnet",
+ "created_at": null,
"context_window": 200000,
- "max_output_tokens": 8192,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -314,8 +426,7 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "batch"
+ "function_calling"
],
"pricing": {
"text_tokens": {
@@ -329,142 +440,227 @@
}
}
},
- "metadata": {}
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "PROVISIONED"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
},
{
- "id": "claude-3-7-sonnet-20250219",
- "name": "Claude Sonnet 3.7",
- "provider": "anthropic",
- "family": "claude-3-7-sonnet",
+ "id": "anthropic.claude-3-5-sonnet-20240620-v1:0:200k",
+ "name": "Claude 3.5 Sonnet",
+ "provider": "bedrock",
+ "family": "claude3_sonnet",
"created_at": null,
"context_window": 200000,
- "max_output_tokens": 64000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 3.0,
- "cached_input_per_million": 3.75,
"output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
- "metadata": {}
- },
- {
- "id": "claude-3-7-sonnet@20250219",
- "name": "Claude Sonnet 3.7",
- "provider": "anthropic",
- "family": "claude-3-7-sonnet",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 64000,
- "knowledge_cutoff": null,
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "PROVISIONED"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
+ },
+ {
+ "id": "anthropic.claude-3-5-sonnet-20240620-v1:0:51k",
+ "name": "Claude 3.5 Sonnet",
+ "provider": "bedrock",
+ "family": "claude3_sonnet",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 4096,
+ "knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 3.0,
- "cached_input_per_million": 3.75,
"output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
- "metadata": {}
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "PROVISIONED"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
},
{
- "id": "claude-3-haiku-20240307",
- "name": "Claude Haiku 3",
- "provider": "anthropic",
- "family": "claude-3-haiku",
+ "id": "anthropic.claude-3-5-sonnet-20241022-v2:0",
+ "name": "Claude 3.5 Sonnet v2",
+ "provider": "bedrock",
+ "family": "claude3_sonnet",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "cached_input_per_million": 0.3,
- "output_per_million": 1.25
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
- "metadata": {}
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "ON_DEMAND"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
},
{
- "id": "claude-3-haiku@20240307",
- "name": "Claude Haiku 3",
- "provider": "anthropic",
- "family": "claude-3-haiku",
+ "id": "anthropic.claude-3-5-sonnet-20241022-v2:0:18k",
+ "name": "Claude 3.5 Sonnet v2",
+ "provider": "bedrock",
+ "family": "claude3_sonnet",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "cached_input_per_million": 0.3,
- "output_per_million": 1.25
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
- "metadata": {}
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "PROVISIONED"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
},
{
- "id": "claude-3-opus-20240229",
- "name": "Claude Opus 3",
- "provider": "anthropic",
- "family": "claude-3-opus",
- "created_at": "2024-02-29 00:00:00 UTC",
+ "id": "anthropic.claude-3-5-sonnet-20241022-v2:0:200k",
+ "name": "Claude 3.5 Sonnet v2",
+ "provider": "bedrock",
+ "family": "claude3_sonnet",
+ "created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
@@ -480,220 +676,240 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "batch"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 75.0
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
},
"batch": {
- "input_per_million": 7.5,
- "output_per_million": 37.5
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
- "metadata": {}
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "PROVISIONED"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
},
{
- "id": "claude-opus-4-1-20250805",
- "name": "Claude Opus 4.1",
- "provider": "anthropic",
- "family": "claude-opus-4-1",
+ "id": "anthropic.claude-3-5-sonnet-20241022-v2:0:51k",
+ "name": "Claude 3.5 Sonnet v2",
+ "provider": "bedrock",
+ "family": "claude3_sonnet",
"created_at": null,
"context_window": 200000,
- "max_output_tokens": 32000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 18.75,
- "output_per_million": 75.0
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
- "metadata": {}
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "PROVISIONED"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
},
{
- "id": "claude-opus-4-1@20250805",
- "name": "Claude Opus 4.1",
- "provider": "anthropic",
- "family": "claude-opus-4-1",
+ "id": "anthropic.claude-3-haiku-20240307-v1:0",
+ "name": "Claude 3 Haiku",
+ "provider": "bedrock",
+ "family": "claude3_haiku",
"created_at": null,
"context_window": 200000,
- "max_output_tokens": 32000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 18.75,
- "output_per_million": 75.0
+ "input_per_million": 0.25,
+ "output_per_million": 1.25
+ },
+ "batch": {
+ "input_per_million": 0.125,
+ "output_per_million": 0.625
}
}
},
- "metadata": {}
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "ON_DEMAND"
+ ],
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
+ ]
+ }
},
{
- "id": "claude-opus-4-20250514",
- "name": "Claude Opus 4",
- "provider": "anthropic",
- "family": "claude-opus-4",
+ "id": "anthropic.claude-3-haiku-20240307-v1:0:200k",
+ "name": "Claude 3 Haiku",
+ "provider": "bedrock",
+ "family": "claude3_haiku",
"created_at": null,
"context_window": 200000,
- "max_output_tokens": 32000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 18.75,
- "output_per_million": 75.0
+ "input_per_million": 0.25,
+ "output_per_million": 1.25
+ },
+ "batch": {
+ "input_per_million": 0.125,
+ "output_per_million": 0.625
}
}
},
- "metadata": {}
- },
- {
- "id": "claude-opus-4@20250514",
- "name": "Claude Opus 4",
- "provider": "anthropic",
- "family": "claude-opus-4",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 32000,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "image",
- "text"
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "PROVISIONED"
],
- "output": [
- "text"
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
]
- },
- "capabilities": [
- "function_calling"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 18.75,
- "output_per_million": 75.0
- }
- }
- },
- "metadata": {}
+ }
},
{
- "id": "claude-sonnet-4-20250514",
- "name": "Claude Sonnet 4",
- "provider": "anthropic",
- "family": "claude-sonnet-4",
+ "id": "anthropic.claude-3-haiku-20240307-v1:0:48k",
+ "name": "Claude 3 Haiku",
+ "provider": "bedrock",
+ "family": "claude3_haiku",
"created_at": null,
"context_window": 200000,
- "max_output_tokens": 64000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "cached_input_per_million": 3.75,
- "output_per_million": 15.0
+ "input_per_million": 0.25,
+ "output_per_million": 1.25
+ },
+ "batch": {
+ "input_per_million": 0.125,
+ "output_per_million": 0.625
}
}
},
- "metadata": {}
- },
- {
- "id": "claude-sonnet-4@20250514",
- "name": "Claude Sonnet 4",
- "provider": "anthropic",
- "family": "claude-sonnet-4",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 64000,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "image",
- "text"
+ "metadata": {
+ "provider_name": "Anthropic",
+ "inference_types": [
+ "PROVISIONED"
],
- "output": [
- "text"
+ "streaming_supported": true,
+ "input_modalities": [
+ "TEXT",
+ "IMAGE"
+ ],
+ "output_modalities": [
+ "TEXT"
]
- },
- "capabilities": [
- "function_calling"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 3.0,
- "cached_input_per_million": 3.75,
- "output_per_million": 15.0
- }
- }
- },
- "metadata": {}
+ }
},
{
- "id": "anthropic.claude-3-5-haiku-20241022-v1:0",
- "name": "Claude 3.5 Haiku",
+ "id": "anthropic.claude-3-opus-20240229-v1:0",
+ "name": "Claude 3 Opus",
"provider": "bedrock",
- "family": "claude3_5_haiku",
+ "family": "claude3_opus",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -715,12 +931,12 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.8,
- "output_per_million": 4.0
+ "input_per_million": 15.0,
+ "output_per_million": 75.0
},
"batch": {
- "input_per_million": 0.4,
- "output_per_million": 2.0
+ "input_per_million": 7.5,
+ "output_per_million": 37.5
}
}
},
@@ -731,7 +947,8 @@
],
"streaming_supported": true,
"input_modalities": [
- "TEXT"
+ "TEXT",
+ "IMAGE"
],
"output_modalities": [
"TEXT"
@@ -739,10 +956,10 @@
}
},
{
- "id": "anthropic.claude-3-5-sonnet-20240620-v1:0",
- "name": "Claude 3.5 Sonnet",
+ "id": "anthropic.claude-3-opus-20240229-v1:0:12k",
+ "name": "Claude 3 Opus",
"provider": "bedrock",
- "family": "claude3_sonnet",
+ "family": "claude3_opus",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -764,21 +981,18 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 15.0,
+ "output_per_million": 75.0
},
"batch": {
- "input_per_million": 1.5,
- "output_per_million": 7.5
+ "input_per_million": 7.5,
+ "output_per_million": 37.5
}
}
},
"metadata": {
"provider_name": "Anthropic",
- "inference_types": [
- "ON_DEMAND",
- "INFERENCE_PROFILE"
- ],
+ "inference_types": [],
"streaming_supported": true,
"input_modalities": [
"TEXT",
@@ -790,10 +1004,10 @@
}
},
{
- "id": "anthropic.claude-3-5-sonnet-20240620-v1:0:18k",
- "name": "Claude 3.5 Sonnet",
+ "id": "anthropic.claude-3-opus-20240229-v1:0:200k",
+ "name": "Claude 3 Opus",
"provider": "bedrock",
- "family": "claude3_sonnet",
+ "family": "claude3_opus",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -815,20 +1029,18 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 15.0,
+ "output_per_million": 75.0
},
"batch": {
- "input_per_million": 1.5,
- "output_per_million": 7.5
+ "input_per_million": 7.5,
+ "output_per_million": 37.5
}
}
},
"metadata": {
"provider_name": "Anthropic",
- "inference_types": [
- "PROVISIONED"
- ],
+ "inference_types": [],
"streaming_supported": true,
"input_modalities": [
"TEXT",
@@ -840,10 +1052,10 @@
}
},
{
- "id": "anthropic.claude-3-5-sonnet-20240620-v1:0:200k",
- "name": "Claude 3.5 Sonnet",
+ "id": "anthropic.claude-3-opus-20240229-v1:0:28k",
+ "name": "Claude 3 Opus",
"provider": "bedrock",
- "family": "claude3_sonnet",
+ "family": "claude3_opus",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -865,20 +1077,18 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 15.0,
+ "output_per_million": 75.0
},
"batch": {
- "input_per_million": 1.5,
- "output_per_million": 7.5
+ "input_per_million": 7.5,
+ "output_per_million": 37.5
}
}
},
"metadata": {
"provider_name": "Anthropic",
- "inference_types": [
- "PROVISIONED"
- ],
+ "inference_types": [],
"streaming_supported": true,
"input_modalities": [
"TEXT",
@@ -890,8 +1100,8 @@
}
},
{
- "id": "anthropic.claude-3-5-sonnet-20240620-v1:0:51k",
- "name": "Claude 3.5 Sonnet",
+ "id": "anthropic.claude-3-sonnet-20240229-v1:0",
+ "name": "Claude 3 Sonnet",
"provider": "bedrock",
"family": "claude3_sonnet",
"created_at": null,
@@ -927,7 +1137,7 @@
"metadata": {
"provider_name": "Anthropic",
"inference_types": [
- "PROVISIONED"
+ "ON_DEMAND"
],
"streaming_supported": true,
"input_modalities": [
@@ -940,8 +1150,8 @@
}
},
{
- "id": "anthropic.claude-3-5-sonnet-20241022-v2:0",
- "name": "Claude 3.5 Sonnet v2",
+ "id": "anthropic.claude-3-sonnet-20240229-v1:0:200k",
+ "name": "Claude 3 Sonnet",
"provider": "bedrock",
"family": "claude3_sonnet",
"created_at": null,
@@ -977,7 +1187,7 @@
"metadata": {
"provider_name": "Anthropic",
"inference_types": [
- "ON_DEMAND"
+ "PROVISIONED"
],
"streaming_supported": true,
"input_modalities": [
@@ -990,8 +1200,8 @@
}
},
{
- "id": "anthropic.claude-3-5-sonnet-20241022-v2:0:18k",
- "name": "Claude 3.5 Sonnet v2",
+ "id": "anthropic.claude-3-sonnet-20240229-v1:0:28k",
+ "name": "Claude 3 Sonnet",
"provider": "bedrock",
"family": "claude3_sonnet",
"created_at": null,
@@ -1040,10 +1250,10 @@
}
},
{
- "id": "anthropic.claude-3-5-sonnet-20241022-v2:0:200k",
- "name": "Claude 3.5 Sonnet v2",
+ "id": "anthropic.claude-instant-v1",
+ "name": "Claude Instant",
"provider": "bedrock",
- "family": "claude3_sonnet",
+ "family": "claude_instant",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1065,24 +1275,23 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 0.8,
+ "output_per_million": 2.4
},
"batch": {
- "input_per_million": 1.5,
- "output_per_million": 7.5
+ "input_per_million": 0.4,
+ "output_per_million": 1.2
}
}
},
"metadata": {
"provider_name": "Anthropic",
"inference_types": [
- "PROVISIONED"
+ "ON_DEMAND"
],
"streaming_supported": true,
"input_modalities": [
- "TEXT",
- "IMAGE"
+ "TEXT"
],
"output_modalities": [
"TEXT"
@@ -1090,10 +1299,10 @@
}
},
{
- "id": "anthropic.claude-3-5-sonnet-20241022-v2:0:51k",
- "name": "Claude 3.5 Sonnet v2",
+ "id": "anthropic.claude-instant-v1:2:100k",
+ "name": "Claude Instant",
"provider": "bedrock",
- "family": "claude3_sonnet",
+ "family": "claude_instant",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1115,12 +1324,12 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 0.8,
+ "output_per_million": 2.4
},
"batch": {
- "input_per_million": 1.5,
- "output_per_million": 7.5
+ "input_per_million": 0.4,
+ "output_per_million": 1.2
}
}
},
@@ -1131,8 +1340,7 @@
],
"streaming_supported": true,
"input_modalities": [
- "TEXT",
- "IMAGE"
+ "TEXT"
],
"output_modalities": [
"TEXT"
@@ -1140,10 +1348,10 @@
}
},
{
- "id": "anthropic.claude-3-haiku-20240307-v1:0",
- "name": "Claude 3 Haiku",
+ "id": "anthropic.claude-v2",
+ "name": "Claude",
"provider": "bedrock",
- "family": "claude3_haiku",
+ "family": "claude2",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1165,12 +1373,12 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "output_per_million": 1.25
+ "input_per_million": 8.0,
+ "output_per_million": 24.0
},
"batch": {
- "input_per_million": 0.125,
- "output_per_million": 0.625
+ "input_per_million": 4.0,
+ "output_per_million": 12.0
}
}
},
@@ -1181,8 +1389,7 @@
],
"streaming_supported": true,
"input_modalities": [
- "TEXT",
- "IMAGE"
+ "TEXT"
],
"output_modalities": [
"TEXT"
@@ -1190,10 +1397,10 @@
}
},
{
- "id": "anthropic.claude-3-haiku-20240307-v1:0:200k",
- "name": "Claude 3 Haiku",
+ "id": "anthropic.claude-v2:0:100k",
+ "name": "Claude",
"provider": "bedrock",
- "family": "claude3_haiku",
+ "family": "claude2",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1215,12 +1422,12 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "output_per_million": 1.25
+ "input_per_million": 8.0,
+ "output_per_million": 24.0
},
"batch": {
- "input_per_million": 0.125,
- "output_per_million": 0.625
+ "input_per_million": 4.0,
+ "output_per_million": 12.0
}
}
},
@@ -1231,8 +1438,7 @@
],
"streaming_supported": true,
"input_modalities": [
- "TEXT",
- "IMAGE"
+ "TEXT"
],
"output_modalities": [
"TEXT"
@@ -1240,10 +1446,10 @@
}
},
{
- "id": "anthropic.claude-3-haiku-20240307-v1:0:48k",
- "name": "Claude 3 Haiku",
+ "id": "anthropic.claude-v2:0:18k",
+ "name": "Claude",
"provider": "bedrock",
- "family": "claude3_haiku",
+ "family": "claude2",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1265,12 +1471,12 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "output_per_million": 1.25
+ "input_per_million": 8.0,
+ "output_per_million": 24.0
},
"batch": {
- "input_per_million": 0.125,
- "output_per_million": 0.625
+ "input_per_million": 4.0,
+ "output_per_million": 12.0
}
}
},
@@ -1281,8 +1487,7 @@
],
"streaming_supported": true,
"input_modalities": [
- "TEXT",
- "IMAGE"
+ "TEXT"
],
"output_modalities": [
"TEXT"
@@ -1290,10 +1495,10 @@
}
},
{
- "id": "anthropic.claude-3-opus-20240229-v1:0",
- "name": "Claude 3 Opus",
+ "id": "anthropic.claude-v2:1",
+ "name": "Claude",
"provider": "bedrock",
- "family": "claude3_opus",
+ "family": "claude2",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1315,12 +1520,12 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 75.0
+ "input_per_million": 8.0,
+ "output_per_million": 24.0
},
"batch": {
- "input_per_million": 7.5,
- "output_per_million": 37.5
+ "input_per_million": 4.0,
+ "output_per_million": 12.0
}
}
},
@@ -1331,8 +1536,7 @@
],
"streaming_supported": true,
"input_modalities": [
- "TEXT",
- "IMAGE"
+ "TEXT"
],
"output_modalities": [
"TEXT"
@@ -1340,10 +1544,10 @@
}
},
{
- "id": "anthropic.claude-3-opus-20240229-v1:0:12k",
- "name": "Claude 3 Opus",
+ "id": "anthropic.claude-v2:1:18k",
+ "name": "Claude",
"provider": "bedrock",
- "family": "claude3_opus",
+ "family": "claude2",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1365,22 +1569,23 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 75.0
+ "input_per_million": 8.0,
+ "output_per_million": 24.0
},
"batch": {
- "input_per_million": 7.5,
- "output_per_million": 37.5
+ "input_per_million": 4.0,
+ "output_per_million": 12.0
}
}
},
"metadata": {
"provider_name": "Anthropic",
- "inference_types": [],
+ "inference_types": [
+ "PROVISIONED"
+ ],
"streaming_supported": true,
"input_modalities": [
- "TEXT",
- "IMAGE"
+ "TEXT"
],
"output_modalities": [
"TEXT"
@@ -1388,10 +1593,10 @@
}
},
{
- "id": "anthropic.claude-3-opus-20240229-v1:0:200k",
- "name": "Claude 3 Opus",
+ "id": "anthropic.claude-v2:1:200k",
+ "name": "Claude",
"provider": "bedrock",
- "family": "claude3_opus",
+ "family": "claude2",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1413,22 +1618,23 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 75.0
+ "input_per_million": 8.0,
+ "output_per_million": 24.0
},
"batch": {
- "input_per_million": 7.5,
- "output_per_million": 37.5
+ "input_per_million": 4.0,
+ "output_per_million": 12.0
}
}
},
"metadata": {
"provider_name": "Anthropic",
- "inference_types": [],
+ "inference_types": [
+ "PROVISIONED"
+ ],
"streaming_supported": true,
"input_modalities": [
- "TEXT",
- "IMAGE"
+ "TEXT"
],
"output_modalities": [
"TEXT"
@@ -1436,10 +1642,10 @@
}
},
{
- "id": "anthropic.claude-3-opus-20240229-v1:0:28k",
- "name": "Claude 3 Opus",
+ "id": "us.anthropic.claude-3-7-sonnet-20250219-v1:0",
+ "name": "Claude 3.7 Sonnet",
"provider": "bedrock",
- "family": "claude3_opus",
+ "family": "claude3_sonnet",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1456,23 +1662,28 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "reasoning",
+ "batch",
+ "citations"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 75.0
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
},
"batch": {
- "input_per_million": 7.5,
- "output_per_million": 37.5
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
"metadata": {
"provider_name": "Anthropic",
- "inference_types": [],
+ "inference_types": [
+ "INFERENCE_PROFILE"
+ ],
"streaming_supported": true,
"input_modalities": [
"TEXT",
@@ -1484,10 +1695,10 @@
}
},
{
- "id": "anthropic.claude-3-sonnet-20240229-v1:0",
- "name": "Claude 3 Sonnet",
+ "id": "us.anthropic.claude-opus-4-1-20250805-v1:0",
+ "name": "Claude Opus 4.1",
"provider": "bedrock",
- "family": "claude3_sonnet",
+ "family": "other",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1509,19 +1720,19 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 0.1,
+ "output_per_million": 0.2
},
"batch": {
- "input_per_million": 1.5,
- "output_per_million": 7.5
+ "input_per_million": 0.05,
+ "output_per_million": 0.1
}
}
},
"metadata": {
"provider_name": "Anthropic",
"inference_types": [
- "ON_DEMAND"
+ "INFERENCE_PROFILE"
],
"streaming_supported": true,
"input_modalities": [
@@ -1534,10 +1745,10 @@
}
},
{
- "id": "anthropic.claude-3-sonnet-20240229-v1:0:200k",
- "name": "Claude 3 Sonnet",
+ "id": "us.anthropic.claude-opus-4-20250514-v1:0",
+ "name": "Claude Opus 4",
"provider": "bedrock",
- "family": "claude3_sonnet",
+ "family": "other",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1559,19 +1770,19 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 0.1,
+ "output_per_million": 0.2
},
"batch": {
- "input_per_million": 1.5,
- "output_per_million": 7.5
+ "input_per_million": 0.05,
+ "output_per_million": 0.1
}
}
},
"metadata": {
"provider_name": "Anthropic",
"inference_types": [
- "PROVISIONED"
+ "INFERENCE_PROFILE"
],
"streaming_supported": true,
"input_modalities": [
@@ -1584,10 +1795,10 @@
}
},
{
- "id": "anthropic.claude-3-sonnet-20240229-v1:0:28k",
- "name": "Claude 3 Sonnet",
+ "id": "us.anthropic.claude-sonnet-4-20250514-v1:0",
+ "name": "Claude Sonnet 4",
"provider": "bedrock",
- "family": "claude3_sonnet",
+ "family": "other",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 4096,
@@ -1609,19 +1820,19 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 0.1,
+ "output_per_million": 0.2
},
"batch": {
- "input_per_million": 1.5,
- "output_per_million": 7.5
+ "input_per_million": 0.05,
+ "output_per_million": 0.1
}
}
},
"metadata": {
"provider_name": "Anthropic",
"inference_types": [
- "PROVISIONED"
+ "INFERENCE_PROFILE"
],
"streaming_supported": true,
"input_modalities": [
@@ -1634,611 +1845,470 @@
}
},
{
- "id": "anthropic.claude-instant-v1",
- "name": "Claude Instant",
- "provider": "bedrock",
- "family": "claude_instant",
+ "id": "deepseek-chat",
+ "name": "deepseek-chat",
+ "provider": "deepseek",
+ "family": "deepseek-chat",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 128000,
+ "max_output_tokens": 8000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.8,
- "output_per_million": 2.4
- },
- "batch": {
- "input_per_million": 0.4,
- "output_per_million": 1.2
+ "input_per_million": 0.27,
+ "cached_input_per_million": 0.07,
+ "output_per_million": 1.1
}
}
},
- "metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "ON_DEMAND"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT"
- ],
- "output_modalities": [
- "TEXT"
- ]
- }
+ "metadata": {}
},
{
- "id": "anthropic.claude-instant-v1:2:100k",
- "name": "Claude Instant",
- "provider": "bedrock",
- "family": "claude_instant",
+ "id": "deepseek-reasoner",
+ "name": "deepseek-reasoner",
+ "provider": "deepseek",
+ "family": "deepseek-reasoner",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": null,
+ "max_output_tokens": 64000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.8,
- "output_per_million": 2.4
- },
- "batch": {
- "input_per_million": 0.4,
- "output_per_million": 1.2
+ "input_per_million": 0.55,
+ "cached_input_per_million": 0.14,
+ "output_per_million": 2.19
}
}
},
- "metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "PROVISIONED"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT"
- ],
- "output_modalities": [
- "TEXT"
- ]
- }
+ "metadata": {}
},
{
- "id": "anthropic.claude-v2",
- "name": "Claude",
- "provider": "bedrock",
- "family": "claude2",
+ "id": "aqa",
+ "name": "Model that performs Attributed Question Answering.",
+ "provider": "gemini",
+ "family": "aqa",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 7168,
+ "max_output_tokens": 1024,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "streaming"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 8.0,
- "output_per_million": 24.0
- },
- "batch": {
- "input_per_million": 4.0,
- "output_per_million": 12.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "ON_DEMAND"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT"
- ],
- "output_modalities": [
- "TEXT"
+ "version": "001",
+ "description": "Model trained to return answers to questions that are grounded in provided sources, along with estimating answerable probability.",
+ "supported_generation_methods": [
+ "generateAnswer"
]
}
},
{
- "id": "anthropic.claude-v2:0:100k",
- "name": "Claude",
- "provider": "bedrock",
- "family": "claude2",
+ "id": "embedding-001",
+ "name": "Embedding 001",
+ "provider": "gemini",
+ "family": "embedding1",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 2048,
+ "max_output_tokens": 1,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "text"
+ "text",
+ "embeddings"
]
},
"capabilities": [
"streaming",
- "function_calling"
+ "batch"
],
"pricing": {
- "text_tokens": {
+ "embeddings": {
"standard": {
- "input_per_million": 8.0,
- "output_per_million": 24.0
- },
- "batch": {
- "input_per_million": 4.0,
- "output_per_million": 12.0
+ "input_per_million": 0.002
}
}
},
"metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "PROVISIONED"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT"
- ],
- "output_modalities": [
- "TEXT"
+ "version": "001",
+ "description": "Obtain a distributed representation of a text.",
+ "supported_generation_methods": [
+ "embedContent"
]
}
},
{
- "id": "anthropic.claude-v2:0:18k",
- "name": "Claude",
- "provider": "bedrock",
- "family": "claude2",
+ "id": "embedding-gecko-001",
+ "name": "Embedding Gecko",
+ "provider": "gemini",
+ "family": "other",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 1024,
+ "max_output_tokens": 1,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "text"
+ "text",
+ "embeddings"
]
},
"capabilities": [
"streaming",
- "function_calling"
+ "batch"
],
"pricing": {
- "text_tokens": {
+ "embeddings": {
"standard": {
- "input_per_million": 8.0,
- "output_per_million": 24.0
- },
- "batch": {
- "input_per_million": 4.0,
- "output_per_million": 12.0
+ "input_per_million": 0.002
}
}
},
"metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "PROVISIONED"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT"
- ],
- "output_modalities": [
- "TEXT"
+ "version": "001",
+ "description": "Obtain a distributed representation of a text.",
+ "supported_generation_methods": [
+ "embedText",
+ "countTextTokens"
]
}
},
{
- "id": "anthropic.claude-v2:1",
- "name": "Claude",
- "provider": "bedrock",
- "family": "claude2",
+ "id": "gemini-1.5-flash",
+ "name": "Gemini 1.5 Flash",
+ "provider": "gemini",
+ "family": "gemini-1.5-flash",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "audio",
"image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 8.0,
- "output_per_million": 24.0
- },
- "batch": {
- "input_per_million": 4.0,
- "output_per_million": 12.0
+ "input_per_million": 0.075,
+ "cached_input_per_million": 0.01875,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "ON_DEMAND"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT"
- ],
- "output_modalities": [
- "TEXT"
+ "version": "001",
+ "description": "Alias that points to the most recent stable version of Gemini 1.5 Flash, our fast and versatile multimodal model for scaling across diverse tasks.",
+ "supported_generation_methods": [
+ "generateContent",
+ "countTokens"
]
}
},
{
- "id": "anthropic.claude-v2:1:18k",
- "name": "Claude",
- "provider": "bedrock",
- "family": "claude2",
+ "id": "gemini-1.5-flash-001",
+ "name": "Gemini 1.5 Flash",
+ "provider": "gemini",
+ "family": "gemini-1.5-flash",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "audio",
"image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 8.0,
- "output_per_million": 24.0
- },
- "batch": {
- "input_per_million": 4.0,
- "output_per_million": 12.0
+ "input_per_million": 0.075,
+ "cached_input_per_million": 0.01875,
+ "output_per_million": 0.3
}
}
},
- "metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "PROVISIONED"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT"
- ],
- "output_modalities": [
- "TEXT"
- ]
- }
+ "metadata": {}
},
{
- "id": "anthropic.claude-v2:1:200k",
- "name": "Claude",
- "provider": "bedrock",
- "family": "claude2",
+ "id": "gemini-1.5-flash-002",
+ "name": "Gemini 1.5 Flash",
+ "provider": "gemini",
+ "family": "gemini-1.5-flash",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "audio",
"image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 8.0,
- "output_per_million": 24.0
- },
- "batch": {
- "input_per_million": 4.0,
- "output_per_million": 12.0
+ "input_per_million": 0.075,
+ "cached_input_per_million": 0.01875,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "PROVISIONED"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT"
- ],
- "output_modalities": [
- "TEXT"
+ "version": "002",
+ "description": "Stable version of Gemini 1.5 Flash, our fast and versatile multimodal model for scaling across diverse tasks, released in September of 2024.",
+ "supported_generation_methods": [
+ "generateContent",
+ "countTokens",
+ "createCachedContent"
]
}
},
{
- "id": "us.anthropic.claude-3-7-sonnet-20250219-v1:0",
- "name": "Claude 3.7 Sonnet",
- "provider": "bedrock",
- "family": "claude3_sonnet",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "id": "gemini-1.5-flash-8b",
+ "name": "Gemini 1.5 Flash-8B",
+ "provider": "gemini",
+ "family": "gemini-1.5-flash-8b",
+ "created_at": null,
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "audio",
"image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
"function_calling",
- "reasoning",
- "batch",
- "citations"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
- },
- "batch": {
- "input_per_million": 1.5,
- "output_per_million": 7.5
+ "input_per_million": 0.075,
+ "cached_input_per_million": 0.01875,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "INFERENCE_PROFILE"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT",
- "IMAGE"
- ],
- "output_modalities": [
- "TEXT"
+ "version": "001",
+ "description": "Stable version of Gemini 1.5 Flash-8B, our smallest and most cost effective Flash model, released in October of 2024.",
+ "supported_generation_methods": [
+ "createCachedContent",
+ "generateContent",
+ "countTokens"
]
}
},
{
- "id": "us.anthropic.claude-opus-4-1-20250805-v1:0",
- "name": "Claude Opus 4.1",
- "provider": "bedrock",
- "family": "other",
+ "id": "gemini-1.5-flash-8b-001",
+ "name": "Gemini 1.5 Flash-8B",
+ "provider": "gemini",
+ "family": "gemini-1.5-flash-8b",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "audio",
"image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "output_per_million": 0.2
- },
- "batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.1
+ "input_per_million": 0.075,
+ "cached_input_per_million": 0.01875,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "INFERENCE_PROFILE"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT",
- "IMAGE"
- ],
- "output_modalities": [
- "TEXT"
+ "version": "001",
+ "description": "Stable version of Gemini 1.5 Flash-8B, our smallest and most cost effective Flash model, released in October of 2024.",
+ "supported_generation_methods": [
+ "createCachedContent",
+ "generateContent",
+ "countTokens"
]
}
},
{
- "id": "us.anthropic.claude-opus-4-20250514-v1:0",
- "name": "Claude Opus 4",
- "provider": "bedrock",
- "family": "other",
+ "id": "gemini-1.5-flash-8b-latest",
+ "name": "Gemini 1.5 Flash-8B",
+ "provider": "gemini",
+ "family": "gemini-1.5-flash-8b",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "audio",
"image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "output_per_million": 0.2
- },
- "batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.1
+ "input_per_million": 0.075,
+ "cached_input_per_million": 0.01875,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "INFERENCE_PROFILE"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT",
- "IMAGE"
- ],
- "output_modalities": [
- "TEXT"
+ "version": "001",
+ "description": "Alias that points to the most recent production (non-experimental) release of Gemini 1.5 Flash-8B, our smallest and most cost effective Flash model, released in October of 2024.",
+ "supported_generation_methods": [
+ "createCachedContent",
+ "generateContent",
+ "countTokens"
]
}
},
{
- "id": "us.anthropic.claude-sonnet-4-20250514-v1:0",
- "name": "Claude Sonnet 4",
- "provider": "bedrock",
- "family": "other",
+ "id": "gemini-1.5-flash-latest",
+ "name": "Gemini 1.5 Flash",
+ "provider": "gemini",
+ "family": "gemini-1.5-flash",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "audio",
"image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "output_per_million": 0.2
- },
- "batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.1
+ "input_per_million": 0.075,
+ "cached_input_per_million": 0.01875,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "provider_name": "Anthropic",
- "inference_types": [
- "INFERENCE_PROFILE"
- ],
- "streaming_supported": true,
- "input_modalities": [
- "TEXT",
- "IMAGE"
- ],
- "output_modalities": [
- "TEXT"
+ "version": "001",
+ "description": "Alias that points to the most recent production (non-experimental) release of Gemini 1.5 Flash, our fast and versatile multimodal model for scaling across diverse tasks.",
+ "supported_generation_methods": [
+ "generateContent",
+ "countTokens"
]
}
},
{
- "id": "deepseek-chat",
- "name": "deepseek-chat",
- "provider": "deepseek",
- "family": "deepseek-chat",
+ "id": "gemini-1.5-pro",
+ "name": "Gemini 1.5 Pro",
+ "provider": "gemini",
+ "family": "gemini-1.5-pro",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 8000,
+ "context_window": 2097152,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
+ "image",
"text"
],
"output": [
@@ -2252,28 +2322,34 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.27,
- "cached_input_per_million": 0.07,
- "output_per_million": 1.1
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.3125,
+ "output_per_million": 5.0
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "deepseek"
+ "version": "001",
+ "description": "Stable version of Gemini 1.5 Pro, our mid-size multimodal model that supports up to 2 million tokens, released in May of 2024.",
+ "supported_generation_methods": [
+ "generateContent",
+ "countTokens"
+ ]
}
},
{
- "id": "deepseek-reasoner",
- "name": "deepseek-reasoner",
- "provider": "deepseek",
- "family": "deepseek-reasoner",
+ "id": "gemini-1.5-pro-001",
+ "name": "Gemini 1.5 Pro",
+ "provider": "gemini",
+ "family": "gemini-1.5-pro",
"created_at": null,
- "context_window": null,
- "max_output_tokens": 64000,
+ "context_window": 2097152,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
+ "image",
"text"
],
"output": [
@@ -2287,28 +2363,27 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.55,
- "cached_input_per_million": 0.14,
- "output_per_million": 2.19
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.3125,
+ "output_per_million": 5.0
}
}
},
- "metadata": {
- "object": "model",
- "owned_by": "deepseek"
- }
+ "metadata": {}
},
{
- "id": "aqa",
- "name": "Model that performs Attributed Question Answering.",
+ "id": "gemini-1.5-pro-002",
+ "name": "Gemini 1.5 Pro",
"provider": "gemini",
- "family": "aqa",
+ "family": "gemini-1.5-pro",
"created_at": null,
- "context_window": 7168,
- "max_output_tokens": 1024,
+ "context_window": 2097152,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
+ "image",
"text"
],
"output": [
@@ -2316,97 +2391,122 @@
]
},
"capabilities": [
- "streaming"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.3125,
+ "output_per_million": 5.0
+ }
+ }
+ },
"metadata": {
- "version": "001",
- "description": "Model trained to return answers to questions that are grounded in provided sources, along with estimating answerable probability.",
+ "version": "002",
+ "description": "Stable version of Gemini 1.5 Pro, our mid-size multimodal model that supports up to 2 million tokens, released in September of 2024.",
"supported_generation_methods": [
- "generateAnswer"
+ "generateContent",
+ "countTokens",
+ "createCachedContent"
]
}
},
{
- "id": "embedding-001",
- "name": "Embedding 001",
+ "id": "gemini-1.5-pro-latest",
+ "name": "Gemini 1.5 Pro",
"provider": "gemini",
- "family": "embedding1",
+ "family": "gemini-1.5-pro",
"created_at": null,
- "context_window": 2048,
- "max_output_tokens": 1,
+ "context_window": 2097152,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
+ "image",
"text"
],
"output": [
- "text",
- "embeddings"
+ "text"
]
},
"capabilities": [
- "streaming",
- "batch"
+ "function_calling",
+ "structured_output"
],
"pricing": {
- "embeddings": {
+ "text_tokens": {
"standard": {
- "input_per_million": 0.002
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.3125,
+ "output_per_million": 5.0
}
}
},
"metadata": {
"version": "001",
- "description": "Obtain a distributed representation of a text.",
+ "description": "Alias that points to the most recent production (non-experimental) release of Gemini 1.5 Pro, our mid-size multimodal model that supports up to 2 million tokens.",
"supported_generation_methods": [
- "embedContent"
+ "generateContent",
+ "countTokens"
]
}
},
{
- "id": "embedding-gecko-001",
- "name": "Embedding Gecko",
+ "id": "gemini-2.0-flash",
+ "name": "Gemini 2.0 Flash",
"provider": "gemini",
- "family": "other",
+ "family": "gemini-2.0-flash",
"created_at": null,
- "context_window": 1024,
- "max_output_tokens": 1,
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
+ "image",
"text"
],
"output": [
- "text",
- "embeddings"
+ "text"
]
},
"capabilities": [
- "streaming",
- "batch"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
- "embeddings": {
+ "text_tokens": {
"standard": {
- "input_per_million": 0.002
+ "input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
"metadata": {
- "version": "001",
- "description": "Obtain a distributed representation of a text.",
+ "version": "2.0",
+ "description": "Gemini 2.0 Flash",
"supported_generation_methods": [
- "embedText",
- "countTextTokens"
+ "generateContent",
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-1.5-flash",
- "name": "Gemini 1.5 Flash",
+ "id": "gemini-2.0-flash-001",
+ "name": "Gemini 2.0 Flash",
"provider": "gemini",
- "family": "gemini-1.5-flash",
+ "family": "gemini-2.0-flash",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 8192,
@@ -2422,32 +2522,39 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "cached_input_per_million": 0.01875,
- "output_per_million": 0.3
+ "input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
"metadata": {
- "version": "001",
- "description": "Alias that points to the most recent stable version of Gemini 1.5 Flash, our fast and versatile multimodal model for scaling across diverse tasks.",
+ "version": "2.0",
+ "description": "Stable version of Gemini 2.0 Flash, our fast and versatile multimodal model for scaling across diverse tasks, released in January of 2025.",
"supported_generation_methods": [
"generateContent",
- "countTokens"
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-1.5-flash-001",
- "name": "Gemini 1.5 Flash",
+ "id": "gemini-2.0-flash-exp",
+ "name": "Gemini 2.0 Flash",
"provider": "gemini",
- "family": "gemini-1.5-flash",
+ "family": "gemini-2.0-flash",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 8192,
@@ -2463,67 +2570,86 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "cached_input_per_million": 0.01875,
- "output_per_million": 0.3
+ "input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
- "metadata": {}
+ "metadata": {
+ "version": "2.0",
+ "description": "Gemini 2.0 Flash Experimental",
+ "supported_generation_methods": [
+ "generateContent",
+ "countTokens",
+ "bidiGenerateContent"
+ ]
+ }
},
{
- "id": "gemini-1.5-flash-002",
- "name": "Gemini 1.5 Flash",
+ "id": "gemini-2.0-flash-exp-image-generation",
+ "name": "Gemini 2.0 Flash (Image Generation) Experimental",
"provider": "gemini",
- "family": "gemini-1.5-flash",
+ "family": "gemini20_flash",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "batch",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "cached_input_per_million": 0.01875,
- "output_per_million": 0.3
+ "input_per_million": 0.1,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
"metadata": {
- "version": "002",
- "description": "Stable version of Gemini 1.5 Flash, our fast and versatile multimodal model for scaling across diverse tasks, released in September of 2024.",
+ "version": "2.0",
+ "description": "Gemini 2.0 Flash (Image Generation) Experimental",
"supported_generation_methods": [
"generateContent",
"countTokens",
- "createCachedContent"
+ "bidiGenerateContent"
]
}
},
{
- "id": "gemini-1.5-flash-8b",
- "name": "Gemini 1.5 Flash-8B",
+ "id": "gemini-2.0-flash-lite",
+ "name": "Gemini 2.0 Flash-Lite",
"provider": "gemini",
- "family": "gemini-1.5-flash-8b",
+ "family": "gemini-2.0-flash-lite",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 8192,
@@ -2539,33 +2665,39 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "cached_input_per_million": 0.01875,
- "output_per_million": 0.3
+ "input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
"metadata": {
- "version": "001",
- "description": "Stable version of Gemini 1.5 Flash-8B, our smallest and most cost effective Flash model, released in October of 2024.",
+ "version": "2.0",
+ "description": "Gemini 2.0 Flash-Lite",
"supported_generation_methods": [
- "createCachedContent",
"generateContent",
- "countTokens"
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-1.5-flash-8b-001",
- "name": "Gemini 1.5 Flash-8B",
+ "id": "gemini-2.0-flash-lite-001",
+ "name": "Gemini 2.0 Flash-Lite",
"provider": "gemini",
- "family": "gemini-1.5-flash-8b",
+ "family": "gemini-2.0-flash-lite",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 8192,
@@ -2581,168 +2713,142 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "cached_input_per_million": 0.01875,
- "output_per_million": 0.3
+ "input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
"metadata": {
- "version": "001",
- "description": "Stable version of Gemini 1.5 Flash-8B, our smallest and most cost effective Flash model, released in October of 2024.",
+ "version": "2.0",
+ "description": "Stable version of Gemini 2.0 Flash-Lite",
"supported_generation_methods": [
- "createCachedContent",
"generateContent",
- "countTokens"
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-1.5-flash-8b-latest",
- "name": "Gemini 1.5 Flash-8B",
+ "id": "gemini-2.0-flash-lite-preview",
+ "name": "Gemini 2.0 Flash-Lite Preview",
"provider": "gemini",
- "family": "gemini-1.5-flash-8b",
+ "family": "gemini20_flash_lite",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "function_calling",
- "structured_output"
+ "streaming",
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.075,
- "cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "001",
- "description": "Alias that points to the most recent production (non-experimental) release of Gemini 1.5 Flash-8B, our smallest and most cost effective Flash model, released in October of 2024.",
+ "version": "preview-02-05",
+ "description": "Preview release (February 5th, 2025) of Gemini 2.0 Flash-Lite",
"supported_generation_methods": [
- "createCachedContent",
"generateContent",
- "countTokens"
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-1.5-flash-latest",
- "name": "Gemini 1.5 Flash",
+ "id": "gemini-2.0-flash-lite-preview-02-05",
+ "name": "Gemini 2.0 Flash-Lite Preview 02-05",
"provider": "gemini",
- "family": "gemini-1.5-flash",
+ "family": "gemini20_flash_lite",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "function_calling",
- "structured_output"
+ "streaming",
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.075,
- "cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "001",
- "description": "Alias that points to the most recent production (non-experimental) release of Gemini 1.5 Flash, our fast and versatile multimodal model for scaling across diverse tasks.",
+ "version": "preview-02-05",
+ "description": "Preview release (February 5th, 2025) of Gemini 2.0 Flash-Lite",
"supported_generation_methods": [
"generateContent",
- "countTokens"
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-1.5-pro",
- "name": "Gemini 1.5 Pro",
+ "id": "gemini-2.0-flash-live-001",
+ "name": "Gemini 2.0 Flash Live",
"provider": "gemini",
- "family": "gemini-1.5-pro",
+ "family": "gemini-2.0-flash-live-001",
"created_at": null,
- "context_window": 2097152,
+ "context_window": 1048576,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
"audio",
- "image",
"text"
],
"output": [
- "text"
- ]
- },
- "capabilities": [
- "function_calling",
- "structured_output"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.3125,
- "output_per_million": 5.0
- }
- }
- },
- "metadata": {
- "version": "001",
- "description": "Stable version of Gemini 1.5 Pro, our mid-size multimodal model that supports up to 2 million tokens, released in May of 2024.",
- "supported_generation_methods": [
- "generateContent",
- "countTokens"
- ]
- }
- },
- {
- "id": "gemini-1.5-pro-001",
- "name": "Gemini 1.5 Pro",
- "provider": "gemini",
- "family": "gemini-1.5-pro",
- "created_at": null,
- "context_window": 2097152,
- "max_output_tokens": 8192,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
"audio",
- "image",
- "text"
- ],
- "output": [
"text"
]
},
@@ -2753,21 +2859,21 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.3125,
- "output_per_million": 5.0
+ "input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 0.4
}
}
},
"metadata": {}
},
{
- "id": "gemini-1.5-pro-002",
- "name": "Gemini 1.5 Pro",
+ "id": "gemini-2.0-flash-preview-image-generation",
+ "name": "Gemini 2.0 Flash Preview Image Generation",
"provider": "gemini",
- "family": "gemini-1.5-pro",
+ "family": "gemini-2.0-flash-preview-image-generation",
"created_at": null,
- "context_window": 2097152,
+ "context_window": 32000,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
@@ -2777,102 +2883,110 @@
"text"
],
"output": [
+ "image",
"text"
]
},
"capabilities": [
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.3125,
- "output_per_million": 5.0
+ "input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 0.4
}
}
},
"metadata": {
- "version": "002",
- "description": "Stable version of Gemini 1.5 Pro, our mid-size multimodal model that supports up to 2 million tokens, released in September of 2024.",
+ "version": "2.0",
+ "description": "Gemini 2.0 Flash Preview Image Generation",
"supported_generation_methods": [
"generateContent",
- "countTokens",
- "createCachedContent"
+ "countTokens"
]
}
},
{
- "id": "gemini-1.5-pro-latest",
- "name": "Gemini 1.5 Pro",
+ "id": "gemini-2.0-flash-thinking-exp",
+ "name": "Gemini 2.5 Flash Preview 05-20",
"provider": "gemini",
- "family": "gemini-1.5-pro",
+ "family": "gemini20_flash",
"created_at": null,
- "context_window": 2097152,
- "max_output_tokens": 8192,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "batch",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.3125,
- "output_per_million": 5.0
+ "input_per_million": 0.1,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
"metadata": {
- "version": "001",
- "description": "Alias that points to the most recent production (non-experimental) release of Gemini 1.5 Pro, our mid-size multimodal model that supports up to 2 million tokens.",
+ "version": "2.5-preview-05-20",
+ "description": "Preview release (April 17th, 2025) of Gemini 2.5 Flash",
"supported_generation_methods": [
"generateContent",
- "countTokens"
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-2.0-flash",
- "name": "Gemini 2.0 Flash",
+ "id": "gemini-2.0-flash-thinking-exp-01-21",
+ "name": "Gemini 2.5 Flash Preview 05-20",
"provider": "gemini",
- "family": "gemini-2.0-flash",
+ "family": "gemini20_flash",
"created_at": null,
"context_window": 1048576,
- "max_output_tokens": 8192,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "batch",
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "batch",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.1,
- "cached_input_per_million": 0.025,
"output_per_million": 0.4
},
"batch": {
@@ -2882,8 +2996,8 @@
}
},
"metadata": {
- "version": "2.0",
- "description": "Gemini 2.0 Flash",
+ "version": "2.5-preview-05-20",
+ "description": "Preview release (April 17th, 2025) of Gemini 2.5 Flash",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -2893,34 +3007,35 @@
}
},
{
- "id": "gemini-2.0-flash-001",
- "name": "Gemini 2.0 Flash",
+ "id": "gemini-2.0-flash-thinking-exp-1219",
+ "name": "Gemini 2.5 Flash Preview 05-20",
"provider": "gemini",
- "family": "gemini-2.0-flash",
+ "family": "gemini20_flash",
"created_at": null,
"context_window": 1048576,
- "max_output_tokens": 8192,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "batch",
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "batch",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.1,
- "cached_input_per_million": 0.025,
"output_per_million": 0.4
},
"batch": {
@@ -2930,8 +3045,8 @@
}
},
"metadata": {
- "version": "2.0",
- "description": "Stable version of Gemini 2.0 Flash, our fast and versatile multimodal model for scaling across diverse tasks, released in January of 2025.",
+ "version": "2.5-preview-05-20",
+ "description": "Preview release (April 17th, 2025) of Gemini 2.5 Flash",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -2941,92 +3056,93 @@
}
},
{
- "id": "gemini-2.0-flash-exp",
- "name": "Gemini 2.0 Flash",
+ "id": "gemini-2.0-pro-exp",
+ "name": "Gemini 2.0 Pro Experimental",
"provider": "gemini",
- "family": "gemini-2.0-flash",
+ "family": "other",
"created_at": null,
"context_window": 1048576,
- "max_output_tokens": 8192,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "batch",
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "cached_input_per_million": 0.025,
- "output_per_million": 0.4
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
},
"batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.2
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "2.0",
- "description": "Gemini 2.0 Flash Experimental",
+ "version": "2.5-exp-03-25",
+ "description": "Experimental release (March 25th, 2025) of Gemini 2.5 Pro",
"supported_generation_methods": [
"generateContent",
"countTokens",
- "bidiGenerateContent"
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-2.0-flash-lite",
- "name": "Gemini 2.0 Flash-Lite",
+ "id": "gemini-2.0-pro-exp-02-05",
+ "name": "Gemini 2.0 Pro Experimental 02-05",
"provider": "gemini",
- "family": "gemini-2.0-flash-lite",
+ "family": "other",
"created_at": null,
"context_window": 1048576,
- "max_output_tokens": 8192,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "batch",
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "cached_input_per_million": 0.025,
- "output_per_million": 0.4
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
},
"batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.2
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "2.0",
- "description": "Gemini 2.0 Flash-Lite",
+ "version": "2.5-exp-03-25",
+ "description": "Experimental release (March 25th, 2025) of Gemini 2.5 Pro",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3036,13 +3152,13 @@
}
},
{
- "id": "gemini-2.0-flash-lite-001",
- "name": "Gemini 2.0 Flash-Lite",
+ "id": "gemini-2.5-flash",
+ "name": "Gemini 2.5 Flash",
"provider": "gemini",
- "family": "gemini-2.0-flash-lite",
+ "family": "gemini-2.5-flash",
"created_at": null,
"context_window": 1048576,
- "max_output_tokens": 8192,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -3062,19 +3178,19 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "cached_input_per_million": 0.025,
- "output_per_million": 0.4
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
},
"batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.2
+ "input_per_million": 0.15,
+ "output_per_million": 1.25
}
}
},
"metadata": {
- "version": "2.0",
- "description": "Stable version of Gemini 2.0 Flash-Lite",
+ "version": "001",
+ "description": "Stable version of Gemini 2.5 Flash, our mid-size multimodal model that supports up to 1 million tokens, released in June of 2025.",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3084,58 +3200,45 @@
}
},
{
- "id": "gemini-2.0-flash-lite-preview",
- "name": "Gemini 2.0 Flash-Lite Preview",
+ "id": "gemini-2.5-flash-exp-native-audio-thinking-dialog",
+ "name": "Gemini 2.5 Flash Native Audio",
"provider": "gemini",
- "family": "gemini20_flash_lite",
+ "family": "gemini-2.5-flash-preview-native-audio-dialog",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 8192,
+ "context_window": 128000,
+ "max_output_tokens": 8000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "audio",
+ "text"
],
"output": [
+ "audio",
"text"
]
},
"capabilities": [
- "streaming",
- "batch"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
- },
- "batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
}
}
},
- "metadata": {
- "version": "preview-02-05",
- "description": "Preview release (February 5th, 2025) of Gemini 2.0 Flash-Lite",
- "supported_generation_methods": [
- "generateContent",
- "countTokens",
- "createCachedContent",
- "batchGenerateContent"
- ]
- }
+ "metadata": {}
},
{
- "id": "gemini-2.0-flash-lite-preview-02-05",
- "name": "Gemini 2.0 Flash-Lite Preview 02-05",
+ "id": "gemini-2.5-flash-image-preview",
+ "name": "Gemini 2.5 Flash Image Preview",
"provider": "gemini",
- "family": "gemini20_flash_lite",
+ "family": "other",
"created_at": null,
- "context_window": 1048576,
+ "context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
@@ -3150,7 +3253,10 @@
},
"capabilities": [
"streaming",
- "batch"
+ "function_calling",
+ "structured_output",
+ "batch",
+ "caching"
],
"pricing": {
"text_tokens": {
@@ -3165,58 +3271,70 @@
}
},
"metadata": {
- "version": "preview-02-05",
- "description": "Preview release (February 5th, 2025) of Gemini 2.0 Flash-Lite",
+ "version": "2.0",
+ "description": "Gemini 2.5 Flash Preview Image",
"supported_generation_methods": [
"generateContent",
- "countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "countTokens"
]
}
},
{
- "id": "gemini-2.0-flash-live-001",
- "name": "Gemini 2.0 Flash Live",
+ "id": "gemini-2.5-flash-lite",
+ "name": "Gemini 2.5 Flash-Lite",
"provider": "gemini",
- "family": "gemini-2.0-flash-live-001",
+ "family": "gemini-2.5-flash-lite",
"created_at": null,
"context_window": 1048576,
- "max_output_tokens": 8192,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
"audio",
+ "image",
"text"
],
"output": [
- "audio",
"text"
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "cached_input_per_million": 0.025,
- "output_per_million": 0.4
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
+ },
+ "batch": {
+ "input_per_million": 0.15,
+ "output_per_million": 1.25
}
}
},
- "metadata": {}
+ "metadata": {
+ "version": "001",
+ "description": "Stable version of Gemini 2.5 Flash-Lite, released in July of 2025",
+ "supported_generation_methods": [
+ "generateContent",
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
+ ]
+ }
},
{
- "id": "gemini-2.0-flash-preview-image-generation",
- "name": "Gemini 2.0 Flash Preview Image Generation",
+ "id": "gemini-2.5-flash-lite-06-17",
+ "name": "Gemini 2.5 Flash-Lite",
"provider": "gemini",
- "family": "gemini-2.0-flash-preview-image-generation",
+ "family": "gemini-2.5-flash-lite",
"created_at": null,
- "context_window": 32000,
- "max_output_tokens": 8192,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -3225,29 +3343,34 @@
"text"
],
"output": [
- "image",
"text"
]
},
"capabilities": [
+ "batch",
+ "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "cached_input_per_million": 0.025,
- "output_per_million": 0.4
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
+ },
+ "batch": {
+ "input_per_million": 0.15,
+ "output_per_million": 1.25
}
}
},
"metadata": {}
},
{
- "id": "gemini-2.0-flash-thinking-exp",
- "name": "Gemini 2.5 Flash Preview 05-20",
+ "id": "gemini-2.5-flash-lite-preview-06-17",
+ "name": "Gemini 2.5 Flash-Lite Preview 06-17",
"provider": "gemini",
- "family": "gemini20_flash",
+ "family": "other",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 65536,
@@ -3264,26 +3387,24 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
- "batch",
- "caching"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "output_per_million": 0.4
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
},
"batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.2
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "2.5-preview-05-20",
- "description": "Preview release (April 17th, 2025) of Gemini 2.5 Flash",
+ "version": "2.5-preview-06-17",
+ "description": "Preview release (June 11th, 2025) of Gemini 2.5 Flash-Lite",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3293,40 +3414,39 @@
}
},
{
- "id": "gemini-2.0-flash-thinking-exp-01-21",
- "name": "Gemini 2.5 Flash Preview 05-20",
+ "id": "gemini-2.5-flash-preview-05-20",
+ "name": "Gemini 2.5 Flash",
"provider": "gemini",
- "family": "gemini20_flash",
+ "family": "gemini-2.5-flash",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "audio",
"image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
"batch",
- "caching"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "output_per_million": 0.4
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
},
"batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.2
+ "input_per_million": 0.15,
+ "output_per_million": 1.25
}
}
},
@@ -3342,94 +3462,114 @@
}
},
{
- "id": "gemini-2.0-flash-thinking-exp-1219",
- "name": "Gemini 2.5 Flash Preview 05-20",
+ "id": "gemini-2.5-flash-preview-native-audio-dialog",
+ "name": "Gemini 2.5 Flash Native Audio",
"provider": "gemini",
- "family": "gemini20_flash",
+ "family": "gemini-2.5-flash-preview-native-audio-dialog",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 128000,
+ "max_output_tokens": 8000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "audio",
+ "text"
],
"output": [
+ "audio",
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch",
- "caching"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "output_per_million": 0.4
- },
- "batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.2
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
+ }
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "id": "gemini-2.5-flash-preview-tts",
+ "name": "Gemini 2.5 Flash Preview Text-to-Speech",
+ "provider": "gemini",
+ "family": "gemini-2.5-flash-preview-tts",
+ "created_at": null,
+ "context_window": 8000,
+ "max_output_tokens": 16000,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "audio"
+ ]
+ },
+ "capabilities": [],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
}
}
},
"metadata": {
- "version": "2.5-preview-05-20",
- "description": "Preview release (April 17th, 2025) of Gemini 2.5 Flash",
+ "version": "gemini-2.5-flash-exp-tts-2025-05-19",
+ "description": "Gemini 2.5 Flash Preview TTS",
"supported_generation_methods": [
- "generateContent",
"countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "generateContent"
]
}
},
{
- "id": "gemini-2.0-pro-exp",
- "name": "Gemini 2.0 Pro Experimental",
+ "id": "gemini-2.5-pro",
+ "name": "Gemini 2.5 Pro",
"provider": "gemini",
- "family": "other",
+ "family": "gemini-2.5-pro",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "audio",
"image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "caching"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.31,
+ "output_per_million": 10.0
},
"batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
+ "input_per_million": 0.625,
+ "output_per_million": 5.0
}
}
},
"metadata": {
- "version": "2.5-exp-03-25",
- "description": "Experimental release (March 25th, 2025) of Gemini 2.5 Pro",
+ "version": "2.5",
+ "description": "Stable release (June 17th, 2025) of Gemini 2.5 Pro",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3439,8 +3579,8 @@
}
},
{
- "id": "gemini-2.0-pro-exp-02-05",
- "name": "Gemini 2.0 Pro Experimental 02-05",
+ "id": "gemini-2.5-pro-preview-03-25",
+ "name": "Gemini 2.5 Pro Preview 03-25",
"provider": "gemini",
"family": "other",
"created_at": null,
@@ -3476,9 +3616,9 @@
}
},
"metadata": {
- "version": "2.5-exp-03-25",
- "description": "Experimental release (March 25th, 2025) of Gemini 2.5 Pro",
- "supported_generation_methods": [
+ "version": "2.5-preview-03-25",
+ "description": "Gemini 2.5 Pro Preview 03-25",
+ "supported_generation_methods": [
"generateContent",
"countTokens",
"createCachedContent",
@@ -3487,45 +3627,45 @@
}
},
{
- "id": "gemini-2.5-flash",
- "name": "Gemini 2.5 Flash",
+ "id": "gemini-2.5-pro-preview-05-06",
+ "name": "Gemini 2.5 Pro Preview 05-06",
"provider": "gemini",
- "family": "gemini-2.5-flash",
+ "family": "other",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "batch",
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
},
"batch": {
- "input_per_million": 0.15,
- "output_per_million": 1.25
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "001",
- "description": "Stable version of Gemini 2.5 Flash, our mid-size multimodal model that supports up to 1 million tokens, released in June of 2025.",
+ "version": "2.5-preview-05-06",
+ "description": "Preview release (May 6th, 2025) of Gemini 2.5 Pro",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3535,133 +3675,136 @@
}
},
{
- "id": "gemini-2.5-flash-exp-native-audio-thinking-dialog",
- "name": "Gemini 2.5 Flash Native Audio",
+ "id": "gemini-2.5-pro-preview-06-05",
+ "name": "Gemini 2.5 Pro Preview",
"provider": "gemini",
- "family": "gemini-2.5-flash-preview-native-audio-dialog",
+ "family": "other",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 8000,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
- "text"
+ "text",
+ "image",
+ "pdf"
],
"output": [
- "audio",
"text"
]
},
"capabilities": [
- "function_calling"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
- "metadata": {}
+ "metadata": {
+ "version": "2.5-preview-06-05",
+ "description": "Preview release (June 5th, 2025) of Gemini 2.5 Pro",
+ "supported_generation_methods": [
+ "generateContent",
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
+ ]
+ }
},
{
- "id": "gemini-2.5-flash-lite",
- "name": "Gemini 2.5 Flash-Lite",
+ "id": "gemini-2.5-pro-preview-tts",
+ "name": "Gemini 2.5 Pro Preview Text-to-Speech",
"provider": "gemini",
- "family": "gemini-2.5-flash-lite",
+ "family": "gemini-2.5-pro-preview-tts",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 8000,
+ "max_output_tokens": 16000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
- "image",
"text"
],
"output": [
- "text"
+ "audio"
]
},
- "capabilities": [
- "batch",
- "function_calling",
- "structured_output"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
- },
- "batch": {
- "input_per_million": 0.15,
- "output_per_million": 1.25
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.31,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "version": "001",
- "description": "Stable verion of Gemini 2.5 Flash-Lite, released in July of 2025",
+ "version": "gemini-2.5-pro-preview-tts-2025-05-19",
+ "description": "Gemini 2.5 Pro Preview TTS",
"supported_generation_methods": [
- "generateContent",
"countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "generateContent"
]
}
},
{
- "id": "gemini-2.5-flash-lite-06-17",
- "name": "Gemini 2.5 Flash-Lite",
+ "id": "gemini-embedding-001",
+ "name": "Gemini Embedding 001",
"provider": "gemini",
- "family": "gemini-2.5-flash-lite",
+ "family": "embedding1",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 2048,
+ "max_output_tokens": 1,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
- "image",
"text"
],
"output": [
- "text"
+ "text",
+ "embeddings"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "streaming",
+ "batch"
],
"pricing": {
- "text_tokens": {
+ "embeddings": {
"standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
- },
- "batch": {
- "input_per_million": 0.15,
- "output_per_million": 1.25
+ "input_per_million": 0.002
}
}
},
- "metadata": {}
+ "metadata": {
+ "version": "001",
+ "description": "Obtain a distributed representation of a text.",
+ "supported_generation_methods": [
+ "embedContent",
+ "countTextTokens",
+ "countTokens"
+ ]
+ }
},
{
- "id": "gemini-2.5-flash-lite-preview-06-17",
- "name": "Gemini 2.5 Flash-Lite Preview 06-17",
+ "id": "gemini-embedding-exp",
+ "name": "Gemini Embedding Experimental",
"provider": "gemini",
- "family": "other",
+ "family": "gemini_embedding_exp",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 8192,
+ "max_output_tokens": 1,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -3670,136 +3813,169 @@
"pdf"
],
"output": [
- "text"
+ "text",
+ "embeddings"
]
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output",
- "batch"
+ "batch",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
+ "input_per_million": 0.002,
+ "output_per_million": 0.004
},
"batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
+ "input_per_million": 0.001,
+ "output_per_million": 0.002
+ }
+ },
+ "embeddings": {
+ "standard": {
+ "input_per_million": 0.002
}
}
},
"metadata": {
- "version": "2.5-preview-06-17",
- "description": "Preview release (June 11th, 2025) of Gemini 2.5 Flash-Lite",
+ "version": "exp-03-07",
+ "description": "Obtain a distributed representation of a text.",
"supported_generation_methods": [
- "generateContent",
- "countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "embedContent",
+ "countTextTokens",
+ "countTokens"
]
}
},
{
- "id": "gemini-2.5-flash-preview-05-20",
- "name": "Gemini 2.5 Flash",
+ "id": "gemini-embedding-exp-03-07",
+ "name": "Gemini Embedding Experimental 03-07",
"provider": "gemini",
- "family": "gemini-2.5-flash",
+ "family": "gemini_embedding_exp",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 8192,
+ "max_output_tokens": 1,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
- "text"
+ "text",
+ "embeddings"
]
},
"capabilities": [
- "batch",
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "batch",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
+ "input_per_million": 0.002,
+ "output_per_million": 0.004
},
"batch": {
- "input_per_million": 0.15,
- "output_per_million": 1.25
+ "input_per_million": 0.001,
+ "output_per_million": 0.002
+ }
+ },
+ "embeddings": {
+ "standard": {
+ "input_per_million": 0.002
}
}
},
"metadata": {
- "version": "2.5-preview-05-20",
- "description": "Preview release (April 17th, 2025) of Gemini 2.5 Flash",
+ "version": "exp-03-07",
+ "description": "Obtain a distributed representation of a text.",
"supported_generation_methods": [
- "generateContent",
- "countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "embedContent",
+ "countTextTokens",
+ "countTokens"
]
}
},
{
- "id": "gemini-2.5-flash-preview-native-audio-dialog",
- "name": "Gemini 2.5 Flash Native Audio",
+ "id": "gemini-exp-1206",
+ "name": "Gemini Experimental 1206",
"provider": "gemini",
- "family": "gemini-2.5-flash-preview-native-audio-dialog",
+ "family": "other",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 8000,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
- "text"
+ "text",
+ "image",
+ "pdf"
],
"output": [
- "audio",
"text"
]
},
"capabilities": [
- "function_calling"
- ],
- "pricing": {
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "caching"
+ ],
+ "pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
- "metadata": {}
+ "metadata": {
+ "version": "2.5-exp-03-25",
+ "description": "Experimental release (March 25th, 2025) of Gemini 2.5 Pro",
+ "supported_generation_methods": [
+ "generateContent",
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
+ ]
+ }
},
{
- "id": "gemini-2.5-flash-preview-tts",
- "name": "Gemini 2.5 Flash Preview TTS",
+ "id": "gemini-live-2.5-flash-preview",
+ "name": "Gemini 2.5 Flash Live",
"provider": "gemini",
- "family": "gemini-2.5-flash-preview-tts",
+ "family": "gemini-live-2.5-flash-preview",
"created_at": null,
- "context_window": 8000,
- "max_output_tokens": 16000,
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
"text"
],
"output": [
- "audio"
+ "audio",
+ "text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "function_calling",
+ "structured_output"
+ ],
"pricing": {
"text_tokens": {
"standard": {
@@ -3809,28 +3985,19 @@
}
}
},
- "metadata": {
- "version": "gemini-2.5-flash-exp-tts-2025-05-19",
- "description": "Gemini 2.5 Flash Preview TTS",
- "supported_generation_methods": [
- "countTokens",
- "generateContent"
- ]
- }
+ "metadata": {}
},
{
- "id": "gemini-2.5-pro",
- "name": "Gemini 2.5 Pro",
+ "id": "gemma-3-12b-it",
+ "name": "Gemma 3 12B",
"provider": "gemini",
- "family": "gemini-2.5-pro",
+ "family": "other",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
- "image",
"text"
],
"output": [
@@ -3838,58 +4005,48 @@
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.31,
- "output_per_million": 10.0
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
},
"batch": {
- "input_per_million": 0.625,
- "output_per_million": 5.0
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "2.5",
- "description": "Stable release (June 17th, 2025) of Gemini 2.5 Pro",
+ "version": "001",
+ "description": null,
"supported_generation_methods": [
"generateContent",
- "countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "countTokens"
]
}
},
{
- "id": "gemini-2.5-pro-preview-03-25",
- "name": "Gemini 2.5 Pro Preview 03-25",
+ "id": "gemma-3-1b-it",
+ "name": "Gemma 3 1B",
"provider": "gemini",
"family": "other",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "caching"
+ "streaming"
],
"pricing": {
"text_tokens": {
@@ -3904,40 +4061,33 @@
}
},
"metadata": {
- "version": "2.5-preview-03-25",
- "description": "Gemini 2.5 Pro Preview 03-25",
+ "version": "001",
+ "description": null,
"supported_generation_methods": [
"generateContent",
- "countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "countTokens"
]
}
},
{
- "id": "gemini-2.5-pro-preview-05-06",
- "name": "Gemini 2.5 Pro Preview 05-06",
+ "id": "gemma-3-27b-it",
+ "name": "Gemma 3 27B",
"provider": "gemini",
"family": "other",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 131072,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "caching"
+ "streaming"
],
"pricing": {
"text_tokens": {
@@ -3952,40 +4102,33 @@
}
},
"metadata": {
- "version": "2.5-preview-05-06",
- "description": "Preview release (May 6th, 2025) of Gemini 2.5 Pro",
+ "version": "001",
+ "description": null,
"supported_generation_methods": [
"generateContent",
- "countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "countTokens"
]
}
},
{
- "id": "gemini-2.5-pro-preview-06-05",
- "name": "Gemini 2.5 Pro Preview",
+ "id": "gemma-3-4b-it",
+ "name": "Gemma 3 4B",
"provider": "gemini",
"family": "other",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "caching"
+ "streaming"
],
"pricing": {
"text_tokens": {
@@ -4000,99 +4143,104 @@
}
},
"metadata": {
- "version": "2.5-preview-06-05",
- "description": "Preview release (June 5th, 2025) of Gemini 2.5 Pro",
+ "version": "001",
+ "description": null,
"supported_generation_methods": [
"generateContent",
- "countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "countTokens"
]
}
},
{
- "id": "gemini-2.5-pro-preview-tts",
- "name": "Gemini 2.5 Pro Preview TTS",
+ "id": "gemma-3n-e2b-it",
+ "name": "Gemma 3n E2B",
"provider": "gemini",
- "family": "gemini-2.5-pro-preview-tts",
+ "family": "other",
"created_at": null,
- "context_window": 8000,
- "max_output_tokens": 16000,
+ "context_window": 8192,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "audio"
+ "text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "streaming"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.31,
- "output_per_million": 10.0
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "gemini-2.5-pro-preview-tts-2025-05-19",
- "description": "Gemini 2.5 Pro Preview TTS",
+ "version": "001",
+ "description": null,
"supported_generation_methods": [
- "countTokens",
- "generateContent"
+ "generateContent",
+ "countTokens"
]
}
},
{
- "id": "gemini-embedding-001",
- "name": "Gemini Embedding 001",
+ "id": "gemma-3n-e4b-it",
+ "name": "Gemma 3n E4B",
"provider": "gemini",
- "family": "embedding1",
+ "family": "other",
"created_at": null,
- "context_window": 2048,
- "max_output_tokens": 1,
+ "context_window": 8192,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "text",
- "embeddings"
+ "text"
]
},
"capabilities": [
- "streaming",
- "batch"
+ "streaming"
],
"pricing": {
- "embeddings": {
+ "text_tokens": {
"standard": {
- "input_per_million": 0.002
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
"version": "001",
- "description": "Obtain a distributed representation of a text.",
+ "description": null,
"supported_generation_methods": [
- "embedContent",
- "countTextTokens",
+ "generateContent",
"countTokens"
]
}
},
{
- "id": "gemini-embedding-exp",
- "name": "Gemini Embedding Experimental",
+ "id": "imagen-3.0-generate-002",
+ "name": "Imagen 3.0",
"provider": "gemini",
- "family": "gemini_embedding_exp",
+ "family": "imagen3",
"created_at": null,
- "context_window": 8192,
- "max_output_tokens": 1,
+ "context_window": 480,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -4101,52 +4249,29 @@
"pdf"
],
"output": [
- "text",
- "embeddings"
+ "image"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch",
- "caching"
+ "streaming"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.002,
- "output_per_million": 0.004
- },
- "batch": {
- "input_per_million": 0.001,
- "output_per_million": 0.002
- }
- },
- "embeddings": {
- "standard": {
- "input_per_million": 0.002
- }
- }
- },
+ "pricing": {},
"metadata": {
- "version": "exp-03-07",
- "description": "Obtain a distributed representation of a text.",
+ "version": "002",
+ "description": "Vertex served Imagen 3.0 002 model",
"supported_generation_methods": [
- "embedContent",
- "countTextTokens",
- "countTokens"
+ "predict"
]
}
},
{
- "id": "gemini-embedding-exp-03-07",
- "name": "Gemini Embedding Experimental 03-07",
+ "id": "learnlm-2.0-flash-experimental",
+ "name": "LearnLM 2.0 Flash Experimental",
"provider": "gemini",
- "family": "gemini_embedding_exp",
+ "family": "other",
"created_at": null,
- "context_window": 8192,
- "max_output_tokens": 1,
+ "context_window": 1048576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -4155,8 +4280,7 @@
"pdf"
],
"output": [
- "text",
- "embeddings"
+ "text"
]
},
"capabilities": [
@@ -4169,118 +4293,97 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.002,
- "output_per_million": 0.004
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
},
"batch": {
- "input_per_million": 0.001,
- "output_per_million": 0.002
- }
- },
- "embeddings": {
- "standard": {
- "input_per_million": 0.002
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "exp-03-07",
- "description": "Obtain a distributed representation of a text.",
+ "version": "2.0",
+ "description": "LearnLM 2.0 Flash Experimental",
"supported_generation_methods": [
- "embedContent",
- "countTextTokens",
+ "generateContent",
"countTokens"
]
}
},
{
- "id": "gemini-exp-1206",
- "name": "Gemini Experimental 1206",
+ "id": "text-embedding-004",
+ "name": "Text Embedding 004",
"provider": "gemini",
- "family": "other",
+ "family": "embedding4",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 2048,
+ "max_output_tokens": 1,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "text"
+ "text",
+ "embeddings"
]
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
- "caching"
+ "batch"
],
"pricing": {
- "text_tokens": {
+ "embeddings": {
"standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
- },
- "batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
+ "input_per_million": 0.002
}
}
},
"metadata": {
- "version": "2.5-exp-03-25",
- "description": "Experimental release (March 25th, 2025) of Gemini 2.5 Pro",
+ "version": "004",
+ "description": "Obtain a distributed representation of a text.",
"supported_generation_methods": [
- "generateContent",
- "countTokens",
- "createCachedContent",
- "batchGenerateContent"
+ "embedContent"
]
}
},
{
- "id": "gemini-live-2.5-flash-preview",
- "name": "Gemini 2.5 Flash Live",
- "provider": "gemini",
- "family": "gemini-live-2.5-flash-preview",
- "created_at": null,
- "context_window": 1048576,
+ "id": "codestral-2411-rc5",
+ "name": "Codestral",
+ "provider": "mistral",
+ "family": "codestral",
+ "created_at": "2024-11-26 00:00:00 +1100",
+ "context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
"text"
],
"output": [
- "audio",
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "batch",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
- }
- }
- },
- "metadata": {}
+ "pricing": {},
+ "metadata": {
+ "object": "model",
+ "owned_by": "mistralai"
+ }
},
{
- "id": "gemma-3-12b-it",
- "name": "Gemma 3 12B",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
+ "id": "codestral-2412",
+ "name": "Codestral",
+ "provider": "mistral",
+ "family": "codestral",
+ "created_at": "2024-12-17 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4293,35 +4396,24 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
- },
- "batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "version": "001",
- "description": null,
- "supported_generation_methods": [
- "generateContent",
- "countTokens"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "gemma-3-1b-it",
- "name": "Gemma 3 1B",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
+ "id": "codestral-2501",
+ "name": "Codestral",
+ "provider": "mistral",
+ "family": "codestral",
+ "created_at": "2025-01-14 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4334,36 +4426,25 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
- },
- "batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "version": "001",
- "description": null,
- "supported_generation_methods": [
- "generateContent",
- "countTokens"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "gemma-3-27b-it",
- "name": "Gemma 3 27B",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
- "context_window": 131072,
+ "id": "codestral-2508",
+ "name": "Codestral",
+ "provider": "mistral",
+ "family": "codestral",
+ "created_at": "2025-08-30 00:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
@@ -4375,35 +4456,24 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
- },
- "batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "version": "001",
- "description": null,
- "supported_generation_methods": [
- "generateContent",
- "countTokens"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "gemma-3-4b-it",
- "name": "Gemma 3 4B",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
+ "id": "codestral-embed",
+ "name": "Codestral",
+ "provider": "mistral",
+ "family": "codestral",
+ "created_at": "2025-05-21 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4412,82 +4482,52 @@
"text"
],
"output": [
- "text"
+ "embeddings"
]
},
"capabilities": [
- "streaming"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
- },
- "batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "version": "001",
- "description": null,
- "supported_generation_methods": [
- "generateContent",
- "countTokens"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "gemma-3n-e2b-it",
- "name": "Gemma 3n E2B",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
- "context_window": 8192,
- "max_output_tokens": 2048,
+ "id": "codestral-embed-2505",
+ "name": "Codestral",
+ "provider": "mistral",
+ "family": "codestral",
+ "created_at": "2025-05-21 00:00:00 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "text"
+ "embeddings"
]
},
"capabilities": [
- "streaming"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
- },
- "batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "version": "001",
- "description": null,
- "supported_generation_methods": [
- "generateContent",
- "countTokens"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "gemma-3n-e4b-it",
- "name": "Gemma 3n E4B",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
- "context_window": 8192,
- "max_output_tokens": 2048,
+ "id": "codestral-latest",
+ "name": "Codestral",
+ "provider": "mistral",
+ "family": "codestral",
+ "created_at": "2025-08-30 00:00:00 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -4498,167 +4538,150 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
- },
- "batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "version": "001",
- "description": null,
- "supported_generation_methods": [
- "generateContent",
- "countTokens"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "imagen-3.0-generate-002",
- "name": "Imagen 3.0",
- "provider": "gemini",
- "family": "imagen3",
- "created_at": null,
- "context_window": 480,
+ "id": "devstral-medium-2507",
+ "name": "Devstral Medium 2507",
+ "provider": "mistral",
+ "family": "mistral",
+ "created_at": "2025-07-09 00:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "image"
+ "text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
- "version": "002",
- "description": "Vertex served Imagen 3.0 002 model",
- "supported_generation_methods": [
- "predict"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "imagen-4.0-generate-001",
- "name": "Imagen 4",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
- "context_window": 480,
+ "id": "devstral-medium-latest",
+ "name": "Devstral Medium Latest",
+ "provider": "mistral",
+ "family": "mistral",
+ "created_at": "2025-07-09 00:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "image"
+ "text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
- "version": "001",
- "description": "Vertex served Imagen 4.0 model",
- "supported_generation_methods": [
- "predict"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "imagen-4.0-generate-preview-06-06",
- "name": "Imagen 4 (Preview)",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
- "context_window": 480,
+ "id": "devstral-small-2505",
+ "name": "Devstral Small 2505",
+ "provider": "mistral",
+ "family": "mistral",
+ "created_at": "2025-05-28 00:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "image"
+ "text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
- "version": "01",
- "description": "Vertex served Imagen 4.0 model",
- "supported_generation_methods": [
- "predict"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "imagen-4.0-ultra-generate-preview-06-06",
- "name": "Imagen 4 Ultra (Preview)",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
- "context_window": 480,
+ "id": "devstral-small-2507",
+ "name": "Devstral Small 2507",
+ "provider": "mistral",
+ "family": "mistral",
+ "created_at": "2025-07-09 00:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "image"
+ "text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
- "version": "01",
- "description": "Vertex served Imagen 4.0 ultra model",
- "supported_generation_methods": [
- "predict"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "learnlm-2.0-flash-experimental",
- "name": "LearnLM 2.0 Flash Experimental",
- "provider": "gemini",
- "family": "other",
- "created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 32768,
+ "id": "devstral-small-latest",
+ "name": "Devstral Small Latest",
+ "provider": "mistral",
+ "family": "mistral",
+ "created_at": "2025-07-09 00:00:00 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
"text"
@@ -4669,72 +4692,50 @@
"function_calling",
"structured_output",
"batch",
- "caching"
+ "fine_tuning"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
- },
- "batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "version": "2.0",
- "description": "LearnLM 2.0 Flash Experimental",
- "supported_generation_methods": [
- "generateContent",
- "countTokens"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "text-embedding-004",
- "name": "Text Embedding 004",
- "provider": "gemini",
- "family": "embedding4",
- "created_at": null,
- "context_window": 2048,
- "max_output_tokens": 1,
+ "id": "magistral-medium-2506",
+ "name": "Magistral Medium 2506",
+ "provider": "mistral",
+ "family": "mistral",
+ "created_at": "2025-06-10 00:00:00 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "text",
- "embeddings"
+ "text"
]
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
+ "reasoning",
"batch"
],
- "pricing": {
- "embeddings": {
- "standard": {
- "input_per_million": 0.002
- }
- }
- },
+ "pricing": {},
"metadata": {
- "version": "004",
- "description": "Obtain a distributed representation of a text.",
- "supported_generation_methods": [
- "embedContent"
- ]
+ "object": "model",
+ "owned_by": "mistralai"
}
},
{
- "id": "codestral-2411-rc5",
- "name": "Codestral",
+ "id": "magistral-medium-2507",
+ "name": "Magistral Medium 2507",
"provider": "mistral",
- "family": "codestral",
- "created_at": "2024-11-26 00:00:00 +0100",
+ "family": "mistral",
+ "created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4750,8 +4751,8 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "predicted_outputs"
+ "reasoning",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -4760,11 +4761,11 @@
}
},
{
- "id": "codestral-2412",
- "name": "Codestral",
+ "id": "magistral-medium-latest",
+ "name": "Magistral Medium Latest",
"provider": "mistral",
- "family": "codestral",
- "created_at": "2024-12-17 00:00:00 +0100",
+ "family": "mistral",
+ "created_at": "2025-06-10 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4780,8 +4781,8 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "predicted_outputs"
+ "reasoning",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -4790,11 +4791,11 @@
}
},
{
- "id": "codestral-2501",
- "name": "Codestral",
+ "id": "magistral-small-2506",
+ "name": "Magistral Small 2506",
"provider": "mistral",
- "family": "codestral",
- "created_at": "2025-01-14 00:00:00 +0100",
+ "family": "mistral",
+ "created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4810,8 +4811,8 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "predicted_outputs"
+ "reasoning",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -4820,11 +4821,11 @@
}
},
{
- "id": "codestral-2508",
- "name": "Codestral",
+ "id": "magistral-small-2507",
+ "name": "Magistral Small 2507",
"provider": "mistral",
- "family": "codestral",
- "created_at": "2025-08-30 00:00:00 +0200",
+ "family": "mistral",
+ "created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4840,8 +4841,8 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "predicted_outputs"
+ "reasoning",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -4850,11 +4851,11 @@
}
},
{
- "id": "codestral-embed",
- "name": "Codestral",
+ "id": "magistral-small-latest",
+ "name": "Magistral Small Latest",
"provider": "mistral",
- "family": "codestral",
- "created_at": "2025-05-21 00:00:00 +0200",
+ "family": "mistral",
+ "created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4863,11 +4864,15 @@
"text"
],
"output": [
- "embeddings"
+ "text"
]
},
"capabilities": [
- "predicted_outputs"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "reasoning",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -4876,11 +4881,11 @@
}
},
{
- "id": "codestral-embed-2505",
- "name": "Codestral",
+ "id": "ministral-3b-2410",
+ "name": "Ministral 3B",
"provider": "mistral",
- "family": "codestral",
- "created_at": "2025-05-21 00:00:00 +0200",
+ "family": "ministral",
+ "created_at": "2024-10-16 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4889,11 +4894,15 @@
"text"
],
"output": [
- "embeddings"
+ "text"
]
},
"capabilities": [
- "predicted_outputs"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "distillation"
],
"pricing": {},
"metadata": {
@@ -4902,11 +4911,11 @@
}
},
{
- "id": "codestral-latest",
- "name": "Codestral",
+ "id": "ministral-3b-latest",
+ "name": "Ministral 3B",
"provider": "mistral",
- "family": "codestral",
- "created_at": "2025-08-30 00:00:00 +0200",
+ "family": "ministral",
+ "created_at": "2024-10-16 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4923,7 +4932,7 @@
"function_calling",
"structured_output",
"batch",
- "predicted_outputs"
+ "distillation"
],
"pricing": {},
"metadata": {
@@ -4932,11 +4941,11 @@
}
},
{
- "id": "devstral-medium-2507",
- "name": "Devstral Medium 2507",
+ "id": "ministral-8b-2410",
+ "name": "Ministral 8B",
"provider": "mistral",
- "family": "mistral",
- "created_at": "2025-07-09 00:00:00 +0200",
+ "family": "ministral",
+ "created_at": "2024-10-16 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4953,7 +4962,7 @@
"function_calling",
"structured_output",
"batch",
- "fine_tuning"
+ "distillation"
],
"pricing": {},
"metadata": {
@@ -4962,11 +4971,11 @@
}
},
{
- "id": "devstral-medium-latest",
- "name": "Devstral Medium Latest",
+ "id": "ministral-8b-latest",
+ "name": "Ministral 8B",
"provider": "mistral",
- "family": "mistral",
- "created_at": "2025-07-09 00:00:00 +0200",
+ "family": "ministral",
+ "created_at": "2024-10-16 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -4983,7 +4992,7 @@
"function_calling",
"structured_output",
"batch",
- "fine_tuning"
+ "distillation"
],
"pricing": {},
"metadata": {
@@ -4992,11 +5001,11 @@
}
},
{
- "id": "devstral-small-2505",
- "name": "Devstral Small 2505",
+ "id": "mistral-embed",
+ "name": "Mistral Embed",
"provider": "mistral",
- "family": "mistral",
- "created_at": "2025-05-28 00:00:00 +0200",
+ "family": "mistral-embed",
+ "created_at": "2024-01-11 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5005,16 +5014,10 @@
"text"
],
"output": [
- "text"
+ "embeddings"
]
},
- "capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch",
- "fine_tuning"
- ],
+ "capabilities": [],
"pricing": {},
"metadata": {
"object": "model",
@@ -5022,11 +5025,11 @@
}
},
{
- "id": "devstral-small-2507",
- "name": "Devstral Small 2507",
+ "id": "mistral-large-2407",
+ "name": "Mistral Large",
"provider": "mistral",
- "family": "mistral",
- "created_at": "2025-07-09 00:00:00 +0200",
+ "family": "mistral-large",
+ "created_at": "2024-07-24 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5052,11 +5055,11 @@
}
},
{
- "id": "devstral-small-latest",
- "name": "Devstral Small Latest",
+ "id": "mistral-large-2411",
+ "name": "Mistral Large",
"provider": "mistral",
- "family": "mistral",
- "created_at": "2025-07-09 00:00:00 +0200",
+ "family": "mistral-large",
+ "created_at": "2024-11-20 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5082,11 +5085,11 @@
}
},
{
- "id": "magistral-medium-2506",
- "name": "Magistral Medium 2506",
+ "id": "mistral-large-latest",
+ "name": "Mistral Large",
"provider": "mistral",
- "family": "mistral",
- "created_at": "2025-06-10 00:00:00 +0200",
+ "family": "mistral-large",
+ "created_at": "2024-11-20 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5102,8 +5105,8 @@
"streaming",
"function_calling",
"structured_output",
- "reasoning",
- "batch"
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
@@ -5112,17 +5115,18 @@
}
},
{
- "id": "magistral-medium-2507",
- "name": "Magistral Medium 2507",
+ "id": "mistral-large-pixtral-2411",
+ "name": "Mistral Large",
"provider": "mistral",
- "family": "mistral",
- "created_at": null,
+ "family": "mistral-large",
+ "created_at": "2024-11-12 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -5132,8 +5136,9 @@
"streaming",
"function_calling",
"structured_output",
- "reasoning",
- "batch"
+ "vision",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
@@ -5142,11 +5147,11 @@
}
},
{
- "id": "magistral-medium-latest",
- "name": "Magistral Medium Latest",
+ "id": "mistral-medium",
+ "name": "Mistral Medium",
"provider": "mistral",
- "family": "mistral",
- "created_at": "2025-06-10 00:00:00 +0200",
+ "family": "mistral-medium",
+ "created_at": "2025-05-06 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5162,8 +5167,9 @@
"streaming",
"function_calling",
"structured_output",
- "reasoning",
- "batch"
+ "vision",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
@@ -5172,11 +5178,11 @@
}
},
{
- "id": "magistral-small-2506",
- "name": "Magistral Small 2506",
+ "id": "mistral-medium-2505",
+ "name": "Mistral Medium",
"provider": "mistral",
- "family": "mistral",
- "created_at": null,
+ "family": "mistral-medium",
+ "created_at": "2025-05-06 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5192,8 +5198,9 @@
"streaming",
"function_calling",
"structured_output",
- "reasoning",
- "batch"
+ "vision",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
@@ -5202,10 +5209,10 @@
}
},
{
- "id": "magistral-small-2507",
- "name": "Magistral Small 2507",
+ "id": "mistral-medium-2508",
+ "name": "Mistral Medium",
"provider": "mistral",
- "family": "mistral",
+ "family": "mistral-medium",
"created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
@@ -5222,8 +5229,9 @@
"streaming",
"function_calling",
"structured_output",
- "reasoning",
- "batch"
+ "vision",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
@@ -5232,11 +5240,11 @@
}
},
{
- "id": "magistral-small-latest",
- "name": "Magistral Small Latest",
+ "id": "mistral-medium-latest",
+ "name": "Mistral Medium",
"provider": "mistral",
- "family": "mistral",
- "created_at": null,
+ "family": "mistral-medium",
+ "created_at": "2025-05-06 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5252,8 +5260,9 @@
"streaming",
"function_calling",
"structured_output",
- "reasoning",
- "batch"
+ "vision",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
@@ -5262,11 +5271,11 @@
}
},
{
- "id": "ministral-3b-2410",
- "name": "Ministral 3B",
+ "id": "mistral-moderation-2411",
+ "name": "Mistral Moderation",
"provider": "mistral",
- "family": "ministral",
- "created_at": "2024-10-16 00:00:00 +0200",
+ "family": "mistral-moderation",
+ "created_at": "2024-11-26 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5279,11 +5288,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch",
- "distillation"
+ "moderation"
],
"pricing": {},
"metadata": {
@@ -5292,11 +5297,11 @@
}
},
{
- "id": "ministral-3b-latest",
- "name": "Ministral 3B",
+ "id": "mistral-moderation-latest",
+ "name": "Mistral Moderation",
"provider": "mistral",
- "family": "ministral",
- "created_at": "2024-10-16 00:00:00 +0200",
+ "family": "mistral-moderation",
+ "created_at": "2024-11-26 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5309,11 +5314,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch",
- "distillation"
+ "moderation"
],
"pricing": {},
"metadata": {
@@ -5322,11 +5323,11 @@
}
},
{
- "id": "ministral-8b-2410",
- "name": "Ministral 8B",
+ "id": "mistral-ocr-2503",
+ "name": "Mistral Ocr 2503",
"provider": "mistral",
- "family": "ministral",
- "created_at": "2024-10-16 00:00:00 +0200",
+ "family": "mistral",
+ "created_at": "2025-03-21 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5339,11 +5340,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch",
- "distillation"
+ "vision"
],
"pricing": {},
"metadata": {
@@ -5352,11 +5349,11 @@
}
},
{
- "id": "ministral-8b-latest",
- "name": "Ministral 8B",
+ "id": "mistral-ocr-2505",
+ "name": "Mistral Ocr 2505",
"provider": "mistral",
- "family": "ministral",
- "created_at": "2024-10-16 00:00:00 +0200",
+ "family": "mistral",
+ "created_at": "2025-05-23 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5369,11 +5366,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch",
- "distillation"
+ "vision"
],
"pricing": {},
"metadata": {
@@ -5382,11 +5375,11 @@
}
},
{
- "id": "mistral-embed",
- "name": "Mistral Embed",
- "provider": "mistral",
- "family": "mistral-embed",
- "created_at": "2024-01-11 00:00:00 +0100",
+ "id": "mistral-ocr-latest",
+ "name": "Mistral Ocr Latest",
+ "provider": "mistral",
+ "family": "mistral",
+ "created_at": "2025-05-23 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5395,10 +5388,12 @@
"text"
],
"output": [
- "embeddings"
+ "text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "vision"
+ ],
"pricing": {},
"metadata": {
"object": "model",
@@ -5406,11 +5401,11 @@
}
},
{
- "id": "mistral-large-2407",
- "name": "Mistral Large",
+ "id": "mistral-saba-2502",
+ "name": "Mistral Saba 2502",
"provider": "mistral",
- "family": "mistral-large",
- "created_at": "2024-07-24 00:00:00 +0200",
+ "family": "mistral",
+ "created_at": "2025-02-18 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5426,8 +5421,7 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "fine_tuning"
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5436,11 +5430,11 @@
}
},
{
- "id": "mistral-large-2411",
- "name": "Mistral Large",
+ "id": "mistral-saba-latest",
+ "name": "Mistral Saba Latest",
"provider": "mistral",
- "family": "mistral-large",
- "created_at": "2024-11-20 00:00:00 +0100",
+ "family": "mistral",
+ "created_at": "2025-02-18 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5456,8 +5450,7 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "fine_tuning"
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5466,11 +5459,11 @@
}
},
{
- "id": "mistral-large-latest",
- "name": "Mistral Large",
+ "id": "mistral-small",
+ "name": "Mistral Small",
"provider": "mistral",
- "family": "mistral-large",
- "created_at": "2024-11-20 00:00:00 +0100",
+ "family": "mistral-small",
+ "created_at": "2023-12-11 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5496,18 +5489,17 @@
}
},
{
- "id": "mistral-large-pixtral-2411",
- "name": "Mistral Large",
+ "id": "mistral-small-2312",
+ "name": "Mistral Small",
"provider": "mistral",
- "family": "mistral-large",
- "created_at": "2024-11-12 00:00:00 +0100",
+ "family": "mistral-small",
+ "created_at": "2023-12-11 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -5515,9 +5507,6 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
- "vision",
"batch",
"fine_tuning"
],
@@ -5528,11 +5517,11 @@
}
},
{
- "id": "mistral-medium",
- "name": "Mistral Medium",
+ "id": "mistral-small-2409",
+ "name": "Mistral Small",
"provider": "mistral",
- "family": "mistral-medium",
- "created_at": "2025-05-06 00:00:00 +0200",
+ "family": "mistral-small",
+ "created_at": "2024-09-18 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5548,7 +5537,6 @@
"streaming",
"function_calling",
"structured_output",
- "vision",
"batch",
"fine_tuning"
],
@@ -5559,11 +5547,11 @@
}
},
{
- "id": "mistral-medium-2505",
- "name": "Mistral Medium",
+ "id": "mistral-small-2501",
+ "name": "Mistral Small",
"provider": "mistral",
- "family": "mistral-medium",
- "created_at": "2025-05-06 00:00:00 +0200",
+ "family": "mistral-small",
+ "created_at": "2025-01-13 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5579,7 +5567,6 @@
"streaming",
"function_calling",
"structured_output",
- "vision",
"batch",
"fine_tuning"
],
@@ -5590,11 +5577,11 @@
}
},
{
- "id": "mistral-medium-2508",
- "name": "Mistral Medium",
+ "id": "mistral-small-2503",
+ "name": "Mistral Small",
"provider": "mistral",
- "family": "mistral-medium",
- "created_at": null,
+ "family": "mistral-small",
+ "created_at": "2025-03-03 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5621,11 +5608,11 @@
}
},
{
- "id": "mistral-medium-latest",
- "name": "Mistral Medium",
+ "id": "mistral-small-2506",
+ "name": "Mistral Small",
"provider": "mistral",
- "family": "mistral-medium",
- "created_at": "2025-05-06 00:00:00 +0200",
+ "family": "mistral-small",
+ "created_at": "2025-06-10 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5652,11 +5639,11 @@
}
},
{
- "id": "mistral-moderation-2411",
- "name": "Mistral Moderation",
+ "id": "mistral-small-latest",
+ "name": "Mistral Small",
"provider": "mistral",
- "family": "mistral-moderation",
- "created_at": "2024-11-26 00:00:00 +0100",
+ "family": "mistral-small",
+ "created_at": "2025-06-10 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5669,7 +5656,11 @@
]
},
"capabilities": [
- "moderation"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch",
+ "fine_tuning"
],
"pricing": {},
"metadata": {
@@ -5678,11 +5669,11 @@
}
},
{
- "id": "mistral-moderation-latest",
- "name": "Mistral Moderation",
+ "id": "mistral-tiny",
+ "name": "Mistral Tiny",
"provider": "mistral",
- "family": "mistral-moderation",
- "created_at": "2024-11-26 00:00:00 +0100",
+ "family": "mistral",
+ "created_at": "2023-09-27 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5695,7 +5686,10 @@
]
},
"capabilities": [
- "moderation"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5704,11 +5698,11 @@
}
},
{
- "id": "mistral-ocr-2503",
- "name": "Mistral Ocr 2503",
+ "id": "mistral-tiny-2312",
+ "name": "Mistral Tiny 2312",
"provider": "mistral",
"family": "mistral",
- "created_at": "2025-03-21 00:00:00 +0100",
+ "created_at": "2023-12-11 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5721,7 +5715,8 @@
]
},
"capabilities": [
- "vision"
+ "streaming",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5730,11 +5725,11 @@
}
},
{
- "id": "mistral-ocr-2505",
- "name": "Mistral Ocr 2505",
+ "id": "mistral-tiny-2407",
+ "name": "Mistral Tiny 2407",
"provider": "mistral",
"family": "mistral",
- "created_at": "2025-05-23 00:00:00 +0200",
+ "created_at": "2024-07-18 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5747,7 +5742,10 @@
]
},
"capabilities": [
- "vision"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5756,11 +5754,11 @@
}
},
{
- "id": "mistral-ocr-latest",
- "name": "Mistral Ocr Latest",
+ "id": "mistral-tiny-latest",
+ "name": "Mistral Tiny Latest",
"provider": "mistral",
"family": "mistral",
- "created_at": "2025-05-23 00:00:00 +0200",
+ "created_at": "2024-07-18 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5773,7 +5771,10 @@
]
},
"capabilities": [
- "vision"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5782,11 +5783,11 @@
}
},
{
- "id": "mistral-saba-2502",
- "name": "Mistral Saba 2502",
+ "id": "open-mistral-7b",
+ "name": "Open Mistral 7b",
"provider": "mistral",
"family": "mistral",
- "created_at": "2025-02-18 00:00:00 +0100",
+ "created_at": "2023-09-27 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5811,11 +5812,11 @@
}
},
{
- "id": "mistral-saba-latest",
- "name": "Mistral Saba Latest",
+ "id": "open-mistral-nemo",
+ "name": "Open Mistral Nemo",
"provider": "mistral",
"family": "mistral",
- "created_at": "2025-02-18 00:00:00 +0100",
+ "created_at": "2024-07-18 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5840,11 +5841,11 @@
}
},
{
- "id": "mistral-small",
- "name": "Mistral Small",
+ "id": "open-mistral-nemo-2407",
+ "name": "Open Mistral Nemo 2407",
"provider": "mistral",
- "family": "mistral-small",
- "created_at": "2023-12-11 00:00:00 +0100",
+ "family": "mistral",
+ "created_at": "2024-07-18 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5860,8 +5861,7 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "fine_tuning"
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5870,11 +5870,11 @@
}
},
{
- "id": "mistral-small-2312",
- "name": "Mistral Small",
+ "id": "open-mixtral-8x22b",
+ "name": "Open Mixtral 8x22b",
"provider": "mistral",
- "family": "mistral-small",
- "created_at": "2023-12-11 00:00:00 +0100",
+ "family": "mistral",
+ "created_at": "2024-04-17 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5888,8 +5888,9 @@
},
"capabilities": [
"streaming",
- "batch",
- "fine_tuning"
+ "function_calling",
+ "structured_output",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5898,11 +5899,11 @@
}
},
{
- "id": "mistral-small-2409",
- "name": "Mistral Small",
+ "id": "open-mixtral-8x22b-2404",
+ "name": "Open Mixtral 8x22b 2404",
"provider": "mistral",
- "family": "mistral-small",
- "created_at": "2024-09-18 00:00:00 +0200",
+ "family": "mistral",
+ "created_at": "2024-04-17 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5918,8 +5919,7 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "fine_tuning"
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5928,11 +5928,11 @@
}
},
{
- "id": "mistral-small-2501",
- "name": "Mistral Small",
+ "id": "open-mixtral-8x7b",
+ "name": "Open Mixtral 8x7b",
"provider": "mistral",
- "family": "mistral-small",
- "created_at": "2025-01-13 00:00:00 +0100",
+ "family": "mistral",
+ "created_at": "2023-12-11 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -5948,8 +5948,7 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "fine_tuning"
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5958,17 +5957,18 @@
}
},
{
- "id": "mistral-small-2503",
- "name": "Mistral Small",
+ "id": "pixtral-12b",
+ "name": "Pixtral 12B",
"provider": "mistral",
- "family": "mistral-small",
- "created_at": "2025-03-03 00:00:00 +0100",
+ "family": "pixtral",
+ "created_at": "2024-09-17 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -5979,8 +5979,7 @@
"function_calling",
"structured_output",
"vision",
- "batch",
- "fine_tuning"
+ "batch"
],
"pricing": {},
"metadata": {
@@ -5989,17 +5988,18 @@
}
},
{
- "id": "mistral-small-2506",
- "name": "Mistral Small",
+ "id": "pixtral-12b-2409",
+ "name": "Pixtral 12B",
"provider": "mistral",
- "family": "mistral-small",
- "created_at": "2025-06-10 00:00:00 +0200",
+ "family": "pixtral",
+ "created_at": "2024-09-17 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -6010,8 +6010,7 @@
"function_calling",
"structured_output",
"vision",
- "batch",
- "fine_tuning"
+ "batch"
],
"pricing": {},
"metadata": {
@@ -6020,17 +6019,18 @@
}
},
{
- "id": "mistral-small-latest",
- "name": "Mistral Small",
+ "id": "pixtral-12b-latest",
+ "name": "Pixtral 12B",
"provider": "mistral",
- "family": "mistral-small",
- "created_at": "2025-06-10 00:00:00 +0200",
+ "family": "pixtral",
+ "created_at": "2024-09-17 00:00:00 +1000",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -6040,8 +6040,8 @@
"streaming",
"function_calling",
"structured_output",
- "batch",
- "fine_tuning"
+ "vision",
+ "batch"
],
"pricing": {},
"metadata": {
@@ -6050,17 +6050,18 @@
}
},
{
- "id": "mistral-tiny",
- "name": "Mistral Tiny",
+ "id": "pixtral-large-2411",
+ "name": "Pixtral Large",
"provider": "mistral",
- "family": "mistral",
- "created_at": "2023-09-27 00:00:00 +0200",
+ "family": "pixtral",
+ "created_at": "2024-11-12 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -6070,6 +6071,7 @@
"streaming",
"function_calling",
"structured_output",
+ "vision",
"batch"
],
"pricing": {},
@@ -6079,17 +6081,18 @@
}
},
{
- "id": "mistral-tiny-2312",
- "name": "Mistral Tiny 2312",
+ "id": "pixtral-large-latest",
+ "name": "Pixtral Large",
"provider": "mistral",
- "family": "mistral",
- "created_at": "2023-12-11 00:00:00 +0100",
+ "family": "pixtral",
+ "created_at": "2024-11-12 00:00:00 +1100",
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -6097,6 +6100,9 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
+ "vision",
"batch"
],
"pricing": {},
@@ -6106,11 +6112,11 @@
}
},
{
- "id": "mistral-tiny-2407",
- "name": "Mistral Tiny 2407",
+ "id": "voxtral-mini-2507",
+ "name": "Voxtral Mini 2507",
"provider": "mistral",
"family": "mistral",
- "created_at": "2024-07-18 00:00:00 +0200",
+ "created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -6123,10 +6129,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch"
+ "streaming"
],
"pricing": {},
"metadata": {
@@ -6135,11 +6138,11 @@
}
},
{
- "id": "mistral-tiny-latest",
- "name": "Mistral Tiny Latest",
+ "id": "voxtral-mini-latest",
+ "name": "Voxtral Mini Latest",
"provider": "mistral",
"family": "mistral",
- "created_at": "2024-07-18 00:00:00 +0200",
+ "created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -6152,10 +6155,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch"
+ "streaming"
],
"pricing": {},
"metadata": {
@@ -6164,11 +6164,11 @@
}
},
{
- "id": "open-mistral-7b",
- "name": "Open Mistral 7b",
+ "id": "voxtral-mini-transcribe-2507",
+ "name": "Voxtral Mini Transcribe 2507",
"provider": "mistral",
"family": "mistral",
- "created_at": "2023-09-27 00:00:00 +0200",
+ "created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -6181,10 +6181,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch"
+ "transcription"
],
"pricing": {},
"metadata": {
@@ -6193,11 +6190,11 @@
}
},
{
- "id": "open-mistral-nemo",
- "name": "Open Mistral Nemo",
+ "id": "voxtral-small-2507",
+ "name": "Voxtral Small 2507",
"provider": "mistral",
"family": "mistral",
- "created_at": "2024-07-18 00:00:00 +0200",
+ "created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -6210,10 +6207,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch"
+ "streaming"
],
"pricing": {},
"metadata": {
@@ -6222,11 +6216,11 @@
}
},
{
- "id": "open-mistral-nemo-2407",
- "name": "Open Mistral Nemo 2407",
+ "id": "voxtral-small-latest",
+ "name": "Voxtral Small Latest",
"provider": "mistral",
"family": "mistral",
- "created_at": "2024-07-18 00:00:00 +0200",
+ "created_at": null,
"context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -6239,10 +6233,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch"
+ "streaming"
],
"pricing": {},
"metadata": {
@@ -6251,281 +6242,315 @@
}
},
{
- "id": "open-mixtral-8x22b",
- "name": "Open Mixtral 8x22b",
- "provider": "mistral",
- "family": "mistral",
- "created_at": "2024-04-17 00:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "babbage-002",
+ "name": "babbage-002",
+ "provider": "openai",
+ "family": "babbage-002",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch"
- ],
- "pricing": {},
+ "capabilities": [],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.4,
+ "output_per_million": 0.4
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "open-mixtral-8x22b-2404",
- "name": "Open Mixtral 8x22b 2404",
- "provider": "mistral",
- "family": "mistral",
- "created_at": "2024-04-17 00:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "chatgpt-4o-latest",
+ "name": "ChatGPT-4o",
+ "provider": "openai",
+ "family": "chatgpt-4o-latest",
+ "created_at": null,
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch"
- ],
- "pricing": {},
+ "capabilities": [],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 5.0,
+ "output_per_million": 15.0
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "open-mixtral-8x7b",
- "name": "Open Mixtral 8x7b",
- "provider": "mistral",
- "family": "mistral",
- "created_at": "2023-12-11 00:00:00 +0100",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "codex-mini-latest",
+ "name": "codex-mini-latest",
+ "provider": "openai",
+ "family": "codex-mini-latest",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
"function_calling",
- "structured_output",
- "batch"
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.5,
+ "cached_input_per_million": 0.375,
+ "output_per_million": 6.0
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "pixtral-12b",
- "name": "Pixtral 12B",
- "provider": "mistral",
- "family": "pixtral",
- "created_at": "2024-09-17 00:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "computer-use-preview",
+ "name": "computer-use-preview",
+ "provider": "openai",
+ "family": "computer-use-preview",
+ "created_at": null,
+ "context_window": 8192,
+ "max_output_tokens": 1024,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "image",
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "vision",
- "batch"
+ "batch",
+ "function_calling"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 12.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 6.0
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "pixtral-12b-2409",
- "name": "Pixtral 12B",
- "provider": "mistral",
- "family": "pixtral",
- "created_at": "2024-09-17 00:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "computer-use-preview-2025-03-11",
+ "name": "computer-use-preview",
+ "provider": "openai",
+ "family": "computer-use-preview",
+ "created_at": null,
+ "context_window": 8192,
+ "max_output_tokens": 1024,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "image",
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "vision",
- "batch"
+ "batch",
+ "function_calling"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 12.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 6.0
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "pixtral-12b-latest",
- "name": "Pixtral 12B",
- "provider": "mistral",
- "family": "pixtral",
- "created_at": "2024-09-17 00:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "dall-e-2",
+ "name": "DALL·E 2",
+ "provider": "openai",
+ "family": "dall-e-2",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
- "text"
+ "embeddings",
+ "image"
]
},
- "capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "vision",
- "batch"
- ],
+ "capabilities": [],
"pricing": {},
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "pixtral-large-2411",
- "name": "Pixtral Large",
- "provider": "mistral",
- "family": "pixtral",
- "created_at": "2024-11-12 00:00:00 +0100",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "dall-e-3",
+ "name": "DALL·E 3",
+ "provider": "openai",
+ "family": "dall-e-3",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
- "text"
+ "embeddings",
+ "image"
]
},
- "capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "vision",
- "batch"
- ],
+ "capabilities": [],
"pricing": {},
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "pixtral-large-latest",
- "name": "Pixtral Large",
- "provider": "mistral",
- "family": "pixtral",
- "created_at": "2024-11-12 00:00:00 +0100",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "davinci-002",
+ "name": "davinci-002",
+ "provider": "openai",
+ "family": "davinci-002",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "vision",
- "batch"
- ],
- "pricing": {},
+ "capabilities": [],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 2.0
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "voxtral-mini-2507",
- "name": "Voxtral Mini 2507",
- "provider": "mistral",
- "family": "mistral",
+ "id": "gpt-3.5-turbo",
+ "name": "GPT-3.5 Turbo",
+ "provider": "openai",
+ "family": "gpt-3.5-turbo",
"created_at": null,
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "context_window": 16385,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
+ },
+ "batch": {
+ "input_per_million": 0.25,
+ "output_per_million": 0.75
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "openai"
}
},
{
- "id": "voxtral-mini-latest",
- "name": "Voxtral Mini Latest",
- "provider": "mistral",
- "family": "mistral",
- "created_at": null,
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "gpt-3.5-turbo-0125",
+ "name": "GPT-3.5 Turbo 0125",
+ "provider": "openai",
+ "family": "gpt35_turbo",
+ "created_at": "2024-01-24 09:19:18 +1100",
+ "context_window": 16385,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -6538,20 +6563,27 @@
"capabilities": [
"streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "voxtral-mini-transcribe-2507",
- "name": "Voxtral Mini Transcribe 2507",
- "provider": "mistral",
- "family": "mistral",
- "created_at": null,
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "gpt-3.5-turbo-1106",
+ "name": "GPT-3.5 Turbo 1106",
+ "provider": "openai",
+ "family": "gpt35_turbo",
+ "created_at": "2023-11-03 08:15:48 +1100",
+ "context_window": 16385,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -6562,22 +6594,29 @@
]
},
"capabilities": [
- "transcription"
+ "streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "voxtral-small-2507",
- "name": "Voxtral Small 2507",
- "provider": "mistral",
- "family": "mistral",
- "created_at": null,
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "gpt-3.5-turbo-16k",
+ "name": "GPT-3.5 Turbo 16k",
+ "provider": "openai",
+ "family": "gpt35_turbo",
+ "created_at": "2023-05-11 08:35:02 +1000",
+ "context_window": 16385,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -6590,20 +6629,27 @@
"capabilities": [
"streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "openai-internal"
}
},
{
- "id": "voxtral-small-latest",
- "name": "Voxtral Small Latest",
- "provider": "mistral",
- "family": "mistral",
- "created_at": null,
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "id": "gpt-3.5-turbo-instruct",
+ "name": "GPT-3.5 Turbo Instruct",
+ "provider": "openai",
+ "family": "gpt35_turbo",
+ "created_at": "2023-08-25 04:23:47 +1000",
+ "context_window": 16385,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -6616,36 +6662,44 @@
"capabilities": [
"streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
+ }
+ }
+ },
"metadata": {
"object": "model",
- "owned_by": "mistralai"
+ "owned_by": "system"
}
},
{
- "id": "babbage-002",
- "name": "babbage-002",
+ "id": "gpt-3.5-turbo-instruct-0914",
+ "name": "GPT-3.5 Turbo Instruct 0914",
"provider": "openai",
- "family": "babbage-002",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": 16384,
+ "family": "gpt35_turbo",
+ "created_at": "2023-09-08 07:34:32 +1000",
+ "context_window": 16385,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "embeddings",
"text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "streaming"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.4,
- "output_per_million": 0.4
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
}
}
},
@@ -6655,17 +6709,16 @@
}
},
{
- "id": "chatgpt-4o-latest",
- "name": "ChatGPT-4o",
+ "id": "gpt-4",
+ "name": "GPT-4",
"provider": "openai",
- "family": "chatgpt-4o-latest",
+ "family": "gpt-4",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "context_window": 8192,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -6673,49 +6726,51 @@
"text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "batch"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 5.0,
- "output_per_million": 15.0
+ "input_per_million": 30.0,
+ "output_per_million": 60.0
+ },
+ "batch": {
+ "input_per_million": 15.0,
+ "output_per_million": 30.0
}
}
},
"metadata": {
"object": "model",
- "owned_by": "system"
+ "owned_by": "openai"
}
},
{
- "id": "codex-mini-latest",
- "name": "codex-mini-latest",
+ "id": "gpt-4-0125-preview",
+ "name": "GPT-4 0125 Preview",
"provider": "openai",
- "family": "codex-mini-latest",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "other",
+ "created_at": "2024-01-24 06:20:12 +1100",
+ "context_window": 4096,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
- "embeddings",
"text"
]
},
"capabilities": [
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.5,
- "cached_input_per_million": 0.375,
- "output_per_million": 6.0
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
}
}
},
@@ -6725,17 +6780,16 @@
}
},
{
- "id": "computer-use-preview",
- "name": "computer-use-preview",
+ "id": "gpt-4-0613",
+ "name": "GPT-4",
"provider": "openai",
- "family": "computer-use-preview",
+ "family": "gpt-4",
"created_at": null,
"context_window": 8192,
- "max_output_tokens": 1024,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -6744,119 +6798,146 @@
]
},
"capabilities": [
- "batch",
- "function_calling"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 12.0
+ "input_per_million": 30.0,
+ "output_per_million": 60.0
},
"batch": {
- "input_per_million": 1.5,
- "output_per_million": 6.0
+ "input_per_million": 15.0,
+ "output_per_million": 30.0
}
}
},
- "metadata": {}
+ "metadata": {
+ "object": "model",
+ "owned_by": "openai"
+ }
},
{
- "id": "computer-use-preview-2025-03-11",
- "name": "computer-use-preview",
+ "id": "gpt-4-1106-preview",
+ "name": "GPT-4 1106 Preview",
"provider": "openai",
- "family": "computer-use-preview",
- "created_at": null,
- "context_window": 8192,
- "max_output_tokens": 1024,
+ "family": "other",
+ "created_at": "2023-11-03 07:33:26 +1100",
+ "context_window": 4096,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
- "embeddings",
"text"
]
},
"capabilities": [
- "batch",
- "function_calling"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 12.0
- },
- "batch": {
- "input_per_million": 1.5,
- "output_per_million": 6.0
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
}
}
},
- "metadata": {}
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
},
{
- "id": "dall-e-2",
- "name": "DALL·E 2",
+ "id": "gpt-4-turbo",
+ "name": "GPT-4 Turbo",
"provider": "openai",
- "family": "dall-e-2",
+ "family": "gpt-4-turbo",
"created_at": null,
- "context_window": null,
- "max_output_tokens": null,
+ "context_window": 128000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
"embeddings",
- "image",
"text"
]
},
- "capabilities": [],
- "pricing": {},
+ "capabilities": [
+ "batch",
+ "function_calling"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 10.0,
+ "output_per_million": 30.0
+ },
+ "batch": {
+ "input_per_million": 5.0,
+ "output_per_million": 15.0
+ }
+ }
+ },
"metadata": {
"object": "model",
"owned_by": "system"
}
},
{
- "id": "dall-e-3",
- "name": "DALL·E 3",
+ "id": "gpt-4-turbo-2024-04-09",
+ "name": "GPT-4 Turbo",
"provider": "openai",
- "family": "dall-e-3",
+ "family": "gpt-4-turbo",
"created_at": null,
- "context_window": null,
- "max_output_tokens": null,
+ "context_window": 128000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
"embeddings",
- "image"
+ "text"
]
},
- "capabilities": [],
- "pricing": {},
+ "capabilities": [
+ "batch",
+ "function_calling"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 10.0,
+ "output_per_million": 30.0
+ },
+ "batch": {
+ "input_per_million": 5.0,
+ "output_per_million": 15.0
+ }
+ }
+ },
"metadata": {
"object": "model",
"owned_by": "system"
}
},
{
- "id": "davinci-002",
- "name": "davinci-002",
+ "id": "gpt-4-turbo-preview",
+ "name": "GPT-4 Turbo Preview",
"provider": "openai",
- "family": "davinci-002",
+ "family": "gpt-4-turbo-preview",
"created_at": null,
- "context_window": null,
- "max_output_tokens": 16384,
+ "context_window": 128000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -6871,8 +6952,8 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 2.0
+ "input_per_million": 10.0,
+ "output_per_million": 30.0
}
}
},
@@ -6882,16 +6963,17 @@
}
},
{
- "id": "gpt-3.5-turbo",
- "name": "GPT-3.5 Turbo",
+ "id": "gpt-4.1",
+ "name": "GPT-4.1",
"provider": "openai",
- "family": "gpt-3.5-turbo",
+ "family": "gpt-4.1",
"created_at": null,
- "context_window": 16385,
- "max_output_tokens": 4096,
+ "context_window": 1047576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -6900,50 +6982,62 @@
]
},
"capabilities": [
- "batch"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
},
"batch": {
- "input_per_million": 0.25,
- "output_per_million": 0.75
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
"metadata": {
"object": "model",
- "owned_by": "openai"
+ "owned_by": "system"
}
},
{
- "id": "gpt-3.5-turbo-0125",
- "name": "GPT-3.5 Turbo 0125",
+ "id": "gpt-4.1-2025-04-14",
+ "name": "GPT-4.1",
"provider": "openai",
- "family": "gpt35_turbo",
- "created_at": "2024-01-23 23:19:18 +0100",
- "context_window": 16385,
- "max_output_tokens": 4096,
+ "family": "gpt-4.1",
+ "created_at": null,
+ "context_window": 1047576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
+ },
+ "batch": {
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
@@ -6953,30 +7047,39 @@
}
},
{
- "id": "gpt-3.5-turbo-1106",
- "name": "GPT-3.5 Turbo 1106",
+ "id": "gpt-4.1-mini",
+ "name": "GPT-4.1 mini",
"provider": "openai",
- "family": "gpt35_turbo",
- "created_at": "2023-11-02 22:15:48 +0100",
- "context_window": 16385,
- "max_output_tokens": 4096,
+ "family": "gpt-4.1-mini",
+ "created_at": null,
+ "context_window": 1047576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 0.4,
+ "cached_input_per_million": 0.1,
+ "output_per_million": 1.6
+ },
+ "batch": {
+ "input_per_million": 0.2,
+ "output_per_million": 0.8
}
}
},
@@ -6986,63 +7089,81 @@
}
},
{
- "id": "gpt-3.5-turbo-16k",
- "name": "GPT-3.5 Turbo 16k",
+ "id": "gpt-4.1-mini-2025-04-14",
+ "name": "GPT-4.1 mini",
"provider": "openai",
- "family": "gpt35_turbo",
- "created_at": "2023-05-11 00:35:02 +0200",
- "context_window": 16385,
- "max_output_tokens": 4096,
+ "family": "gpt-4.1-mini",
+ "created_at": null,
+ "context_window": 1047576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 0.4,
+ "cached_input_per_million": 0.1,
+ "output_per_million": 1.6
+ },
+ "batch": {
+ "input_per_million": 0.2,
+ "output_per_million": 0.8
}
}
},
"metadata": {
"object": "model",
- "owned_by": "openai-internal"
+ "owned_by": "system"
}
},
{
- "id": "gpt-3.5-turbo-instruct",
- "name": "GPT-3.5 Turbo Instruct",
+ "id": "gpt-4.1-nano",
+ "name": "GPT-4.1 nano",
"provider": "openai",
- "family": "gpt35_turbo",
- "created_at": "2023-08-24 20:23:47 +0200",
- "context_window": 16385,
- "max_output_tokens": 4096,
+ "family": "gpt-4.1-nano",
+ "created_at": null,
+ "context_window": 1047576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
@@ -7052,49 +7173,59 @@
}
},
{
- "id": "gpt-3.5-turbo-instruct-0914",
- "name": "GPT-3.5 Turbo Instruct 0914",
+ "id": "gpt-4.1-nano-2025-04-14",
+ "name": "GPT-4.1 nano",
"provider": "openai",
- "family": "gpt35_turbo",
- "created_at": "2023-09-07 23:34:32 +0200",
- "context_window": 16385,
- "max_output_tokens": 4096,
+ "family": "gpt-4.1-nano",
+ "created_at": null,
+ "context_window": 1047576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
- }
- }
- },
+ "input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
+ }
+ }
+ },
"metadata": {
"object": "model",
"owned_by": "system"
}
},
{
- "id": "gpt-4",
- "name": "GPT-4",
+ "id": "gpt-4.5-preview",
+ "name": "GPT-4.5 Preview (Deprecated)",
"provider": "openai",
- "family": "gpt-4",
+ "family": "gpt-4.5-preview",
"created_at": null,
- "context_window": 8192,
- "max_output_tokens": 8192,
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -7103,69 +7234,76 @@
]
},
"capabilities": [
- "batch"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 30.0,
- "output_per_million": 60.0
+ "input_per_million": 75.0,
+ "cached_input_per_million": 37.5,
+ "output_per_million": 150.0
},
"batch": {
- "input_per_million": 15.0,
- "output_per_million": 30.0
+ "input_per_million": 37.5,
+ "output_per_million": 75.0
}
}
},
- "metadata": {
- "object": "model",
- "owned_by": "openai"
- }
+ "metadata": {}
},
{
- "id": "gpt-4-0125-preview",
- "name": "GPT-4 0125 Preview",
+ "id": "gpt-4.5-preview-2025-02-27",
+ "name": "GPT-4.5 Preview (Deprecated)",
"provider": "openai",
- "family": "other",
- "created_at": "2024-01-23 20:20:12 +0100",
- "context_window": 4096,
+ "family": "gpt-4.5-preview",
+ "created_at": null,
+ "context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 75.0,
+ "cached_input_per_million": 37.5,
+ "output_per_million": 150.0
+ },
+ "batch": {
+ "input_per_million": 37.5,
+ "output_per_million": 75.0
}
}
},
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
+ "metadata": {}
},
{
- "id": "gpt-4-0613",
- "name": "GPT-4",
+ "id": "gpt-4o",
+ "name": "GPT-4o",
"provider": "openai",
- "family": "gpt-4",
+ "family": "gpt-4o",
"created_at": null,
- "context_window": 8192,
- "max_output_tokens": 8192,
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -7174,50 +7312,57 @@
]
},
"capabilities": [
- "batch"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 30.0,
- "output_per_million": 60.0
+ "input_per_million": 2.5,
+ "cached_input_per_million": 1.25,
+ "output_per_million": 10.0
},
"batch": {
- "input_per_million": 15.0,
- "output_per_million": 30.0
+ "input_per_million": 1.25,
+ "output_per_million": 5.0
}
}
},
"metadata": {
"object": "model",
- "owned_by": "openai"
+ "owned_by": "system"
}
},
{
- "id": "gpt-4-1106-preview",
- "name": "GPT-4 1106 Preview",
+ "id": "gpt-4o-2024-05-13",
+ "name": "GPT-4o 20240513",
"provider": "openai",
- "family": "other",
- "created_at": "2023-11-02 21:33:26 +0100",
- "context_window": 4096,
+ "family": "gpt4o",
+ "created_at": "2024-05-11 05:08:52 +1000",
+ "context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -7227,13 +7372,13 @@
}
},
{
- "id": "gpt-4-turbo",
- "name": "GPT-4 Turbo",
+ "id": "gpt-4o-2024-08-06",
+ "name": "GPT-4o",
"provider": "openai",
- "family": "gpt-4-turbo",
+ "family": "gpt-4o",
"created_at": null,
"context_window": 128000,
- "max_output_tokens": 4096,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -7247,17 +7392,19 @@
},
"capabilities": [
"batch",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 10.0,
- "output_per_million": 30.0
+ "input_per_million": 2.5,
+ "cached_input_per_million": 1.25,
+ "output_per_million": 10.0
},
"batch": {
- "input_per_million": 5.0,
- "output_per_million": 15.0
+ "input_per_million": 1.25,
+ "output_per_million": 5.0
}
}
},
@@ -7267,37 +7414,34 @@
}
},
{
- "id": "gpt-4-turbo-2024-04-09",
- "name": "GPT-4 Turbo",
+ "id": "gpt-4o-2024-11-20",
+ "name": "GPT-4o 20241120",
"provider": "openai",
- "family": "gpt-4-turbo",
- "created_at": null,
+ "family": "gpt4o",
+ "created_at": "2025-02-12 14:39:03 +1100",
"context_window": 128000,
- "max_output_tokens": 4096,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
- "embeddings",
"text"
]
},
"capabilities": [
- "batch",
- "function_calling"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 10.0,
- "output_per_million": 30.0
- },
- "batch": {
- "input_per_million": 5.0,
- "output_per_million": 15.0
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -7307,29 +7451,33 @@
}
},
{
- "id": "gpt-4-turbo-preview",
- "name": "GPT-4 Turbo Preview",
+ "id": "gpt-4o-audio-preview",
+ "name": "GPT-4o Audio",
"provider": "openai",
- "family": "gpt-4-turbo-preview",
+ "family": "gpt-4o-audio-preview",
"created_at": null,
"context_window": 128000,
- "max_output_tokens": 4096,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
"text"
],
"output": [
+ "audio",
"embeddings",
"text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "function_calling"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 10.0,
- "output_per_million": 30.0
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -7339,39 +7487,34 @@
}
},
{
- "id": "gpt-4.1",
- "name": "GPT-4.1",
+ "id": "gpt-4o-audio-preview-2024-10-01",
+ "name": "GPT-4o-Audio Preview 20241001",
"provider": "openai",
- "family": "gpt-4.1",
- "created_at": null,
- "context_window": 1047576,
- "max_output_tokens": 32768,
+ "family": "gpt4o_audio",
+ "created_at": "2024-09-27 08:17:22 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
- "text"
+ "text",
+ "audio"
],
"output": [
- "embeddings",
- "text"
+ "text",
+ "audio"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "streaming",
+ "speech_generation",
+ "transcription"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "cached_input_per_million": 0.5,
- "output_per_million": 8.0
- },
- "batch": {
- "input_per_million": 1.0,
- "output_per_million": 4.0
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -7381,39 +7524,34 @@
}
},
{
- "id": "gpt-4.1-2025-04-14",
- "name": "GPT-4.1",
+ "id": "gpt-4o-audio-preview-2024-12-17",
+ "name": "GPT-4o-Audio Preview 20241217",
"provider": "openai",
- "family": "gpt-4.1",
- "created_at": null,
- "context_window": 1047576,
- "max_output_tokens": 32768,
+ "family": "gpt4o_audio",
+ "created_at": "2024-12-13 07:10:39 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
- "text"
+ "text",
+ "audio"
],
"output": [
- "embeddings",
- "text"
+ "text",
+ "audio"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "streaming",
+ "speech_generation",
+ "transcription"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "cached_input_per_million": 0.5,
- "output_per_million": 8.0
- },
- "batch": {
- "input_per_million": 1.0,
- "output_per_million": 4.0
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -7423,39 +7561,34 @@
}
},
{
- "id": "gpt-4.1-mini",
- "name": "GPT-4.1 mini",
+ "id": "gpt-4o-audio-preview-2025-06-03",
+ "name": "GPT-4o-Audio Preview 20250603",
"provider": "openai",
- "family": "gpt-4.1-mini",
- "created_at": null,
- "context_window": 1047576,
- "max_output_tokens": 32768,
- "knowledge_cutoff": null,
- "modalities": {
+ "family": "gpt4o_audio",
+ "created_at": "2025-06-03 09:54:58 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
+ "knowledge_cutoff": null,
+ "modalities": {
"input": [
- "image",
- "text"
+ "text",
+ "audio"
],
"output": [
- "embeddings",
- "text"
+ "text",
+ "audio"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "streaming",
+ "speech_generation",
+ "transcription"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.4,
- "cached_input_per_million": 0.1,
- "output_per_million": 1.6
- },
- "batch": {
- "input_per_million": 0.2,
- "output_per_million": 0.8
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -7465,13 +7598,13 @@
}
},
{
- "id": "gpt-4.1-mini-2025-04-14",
- "name": "GPT-4.1 mini",
+ "id": "gpt-4o-mini",
+ "name": "GPT-4o mini",
"provider": "openai",
- "family": "gpt-4.1-mini",
+ "family": "gpt-4o-mini",
"created_at": null,
- "context_window": 1047576,
- "max_output_tokens": 32768,
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -7491,13 +7624,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.4,
- "cached_input_per_million": 0.1,
- "output_per_million": 1.6
+ "input_per_million": 0.15,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 0.6
},
"batch": {
- "input_per_million": 0.2,
- "output_per_million": 0.8
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
}
}
},
@@ -7507,13 +7640,13 @@
}
},
{
- "id": "gpt-4.1-nano",
- "name": "GPT-4.1 nano",
+ "id": "gpt-4o-mini-2024-07-18",
+ "name": "GPT-4o mini",
"provider": "openai",
- "family": "gpt-4.1-nano",
+ "family": "gpt-4o-mini",
"created_at": null,
- "context_window": 1047576,
- "max_output_tokens": 32768,
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -7533,13 +7666,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "cached_input_per_million": 0.025,
- "output_per_million": 0.4
+ "input_per_million": 0.15,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 0.6
},
"batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.2
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
}
}
},
@@ -7549,39 +7682,33 @@
}
},
{
- "id": "gpt-4.1-nano-2025-04-14",
- "name": "GPT-4.1 nano",
+ "id": "gpt-4o-mini-audio-preview",
+ "name": "GPT-4o mini Audio",
"provider": "openai",
- "family": "gpt-4.1-nano",
+ "family": "gpt-4o-mini-audio-preview",
"created_at": null,
- "context_window": 1047576,
- "max_output_tokens": 32768,
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
+ "audio",
"text"
],
"output": [
+ "audio",
"embeddings",
"text"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "cached_input_per_million": 0.025,
- "output_per_million": 0.4
- },
- "batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.2
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
}
}
},
@@ -7591,117 +7718,107 @@
}
},
{
- "id": "gpt-4.5-preview",
- "name": "GPT-4.5 Preview (Deprecated)",
+ "id": "gpt-4o-mini-audio-preview-2024-12-17",
+ "name": "GPT-4o mini Audio",
"provider": "openai",
- "family": "gpt-4.5-preview",
+ "family": "gpt-4o-mini-audio-preview",
"created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
+ "audio",
"text"
],
"output": [
+ "audio",
"embeddings",
"text"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 75.0,
- "cached_input_per_million": 37.5,
- "output_per_million": 150.0
- },
- "batch": {
- "input_per_million": 37.5,
- "output_per_million": 75.0
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
}
}
},
- "metadata": {}
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
},
{
- "id": "gpt-4.5-preview-2025-02-27",
- "name": "GPT-4.5 Preview (Deprecated)",
+ "id": "gpt-4o-mini-realtime-preview",
+ "name": "GPT-4o mini Realtime",
"provider": "openai",
- "family": "gpt-4.5-preview",
+ "family": "gpt-4o-mini-realtime-preview",
"created_at": null,
"context_window": 128000,
- "max_output_tokens": 16384,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
+ "audio",
"text"
],
"output": [
+ "audio",
"embeddings",
"text"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 75.0,
- "cached_input_per_million": 37.5,
- "output_per_million": 150.0
- },
- "batch": {
- "input_per_million": 37.5,
- "output_per_million": 75.0
+ "input_per_million": 0.6,
+ "cached_input_per_million": 0.3,
+ "output_per_million": 2.4
}
}
},
- "metadata": {}
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
},
{
- "id": "gpt-4o",
- "name": "GPT-4o",
+ "id": "gpt-4o-mini-realtime-preview-2024-12-17",
+ "name": "GPT-4o mini Realtime",
"provider": "openai",
- "family": "gpt-4o",
+ "family": "gpt-4o-mini-realtime-preview",
"created_at": null,
"context_window": 128000,
- "max_output_tokens": 16384,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
+ "audio",
"text"
],
"output": [
+ "audio",
"embeddings",
"text"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "cached_input_per_million": 1.25,
- "output_per_million": 10.0
- },
- "batch": {
- "input_per_million": 1.25,
- "output_per_million": 5.0
+ "input_per_million": 0.6,
+ "cached_input_per_million": 0.3,
+ "output_per_million": 2.4
}
}
},
@@ -7711,34 +7828,31 @@
}
},
{
- "id": "gpt-4o-2024-05-13",
- "name": "GPT-4o 20240513",
+ "id": "gpt-4o-mini-search-preview",
+ "name": "GPT-4o mini Search Preview",
"provider": "openai",
- "family": "gpt4o",
- "created_at": "2024-05-10 21:08:52 +0200",
+ "family": "gpt-4o-mini-search-preview",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
}
}
},
@@ -7748,17 +7862,16 @@
}
},
{
- "id": "gpt-4o-2024-08-06",
- "name": "GPT-4o",
+ "id": "gpt-4o-mini-search-preview-2025-03-11",
+ "name": "GPT-4o mini Search Preview",
"provider": "openai",
- "family": "gpt-4o",
+ "family": "gpt-4o-mini-search-preview",
"created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -7767,20 +7880,13 @@
]
},
"capabilities": [
- "batch",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "cached_input_per_million": 1.25,
- "output_per_million": 10.0
- },
- "batch": {
- "input_per_million": 1.25,
- "output_per_million": 5.0
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
}
}
},
@@ -7790,34 +7896,30 @@
}
},
{
- "id": "gpt-4o-2024-11-20",
- "name": "GPT-4o 20241120",
+ "id": "gpt-4o-mini-transcribe",
+ "name": "GPT-4o mini Transcribe",
"provider": "openai",
- "family": "gpt4o",
- "created_at": "2025-02-12 04:39:03 +0100",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "family": "gpt-4o-mini-transcribe",
+ "created_at": null,
+ "context_window": 16000,
+ "max_output_tokens": 2000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "audio",
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 1.25,
+ "output_per_million": 5.0
}
}
},
@@ -7827,33 +7929,29 @@
}
},
{
- "id": "gpt-4o-audio-preview",
- "name": "GPT-4o Audio",
+ "id": "gpt-4o-mini-tts",
+ "name": "GPT-4o mini TTS",
"provider": "openai",
- "family": "gpt-4o-audio-preview",
+ "family": "gpt-4o-mini-tts",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "context_window": 2000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
"text"
],
"output": [
"audio",
- "embeddings",
- "text"
+ "embeddings"
]
},
- "capabilities": [
- "function_calling"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 0.6,
+ "output_per_million": 12.0
}
}
},
@@ -7863,34 +7961,34 @@
}
},
{
- "id": "gpt-4o-audio-preview-2024-10-01",
- "name": "GPT-4o-Audio Preview 20241001",
+ "id": "gpt-4o-realtime-preview",
+ "name": "GPT-4o Realtime",
"provider": "openai",
- "family": "gpt4o_audio",
- "created_at": "2024-09-27 00:17:22 +0200",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "family": "gpt-4o-realtime-preview",
+ "created_at": null,
+ "context_window": 32000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "audio",
+ "text"
],
"output": [
- "text",
- "audio"
+ "audio",
+ "embeddings",
+ "text"
]
},
"capabilities": [
- "streaming",
- "speech_generation",
- "transcription"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 5.0,
+ "cached_input_per_million": 2.5,
+ "output_per_million": 20.0
}
}
},
@@ -7900,34 +7998,30 @@
}
},
{
- "id": "gpt-4o-audio-preview-2024-12-17",
- "name": "GPT-4o-Audio Preview 20241217",
+ "id": "gpt-4o-realtime-preview-2024-10-01",
+ "name": "GPT-4o-Realtime Preview 20241001",
"provider": "openai",
- "family": "gpt4o_audio",
- "created_at": "2024-12-12 21:10:39 +0100",
+ "family": "gpt4o_realtime",
+ "created_at": "2024-09-24 08:49:26 +1000",
"context_window": 128000,
- "max_output_tokens": 16384,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "text"
],
"output": [
- "text",
- "audio"
+ "text"
]
},
"capabilities": [
- "streaming",
- "speech_generation",
- "transcription"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 5.0,
+ "output_per_million": 20.0
}
}
},
@@ -7937,34 +8031,30 @@
}
},
{
- "id": "gpt-4o-audio-preview-2025-06-03",
- "name": "GPT-4o-Audio Preview 20250603",
+ "id": "gpt-4o-realtime-preview-2024-12-17",
+ "name": "GPT-4o-Realtime Preview 20241217",
"provider": "openai",
- "family": "gpt4o_audio",
- "created_at": "2025-06-03 01:54:58 +0200",
+ "family": "gpt4o_realtime",
+ "created_at": "2024-12-12 06:30:30 +1100",
"context_window": 128000,
- "max_output_tokens": 16384,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "text"
],
"output": [
- "text",
- "audio"
+ "text"
]
},
"capabilities": [
- "streaming",
- "speech_generation",
- "transcription"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 5.0,
+ "output_per_million": 20.0
}
}
},
@@ -7974,39 +8064,30 @@
}
},
{
- "id": "gpt-4o-mini",
- "name": "GPT-4o mini",
+ "id": "gpt-4o-realtime-preview-2025-06-03",
+ "name": "GPT-4o-Realtime Preview 20250603",
"provider": "openai",
- "family": "gpt-4o-mini",
- "created_at": null,
+ "family": "gpt4o_realtime",
+ "created_at": "2025-06-03 09:43:58 +1000",
"context_window": 128000,
- "max_output_tokens": 16384,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
- "embeddings",
"text"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "cached_input_per_million": 0.075,
- "output_per_million": 0.6
- },
- "batch": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
+ "input_per_million": 5.0,
+ "output_per_million": 20.0
}
}
},
@@ -8016,17 +8097,16 @@
}
},
{
- "id": "gpt-4o-mini-2024-07-18",
- "name": "GPT-4o mini",
+ "id": "gpt-4o-search-preview",
+ "name": "GPT-4o Search Preview",
"provider": "openai",
- "family": "gpt-4o-mini",
+ "family": "gpt-4o-search-preview",
"created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -8035,20 +8115,13 @@
]
},
"capabilities": [
- "batch",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "cached_input_per_million": 0.075,
- "output_per_million": 0.6
- },
- "batch": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -8058,33 +8131,31 @@
}
},
{
- "id": "gpt-4o-mini-audio-preview",
- "name": "GPT-4o mini Audio",
+ "id": "gpt-4o-search-preview-2025-03-11",
+ "name": "GPT-4o Search Preview",
"provider": "openai",
- "family": "gpt-4o-mini-audio-preview",
+ "family": "gpt-4o-search-preview",
"created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
"text"
],
"output": [
- "audio",
"embeddings",
"text"
]
},
"capabilities": [
- "function_calling"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -8094,34 +8165,30 @@
}
},
{
- "id": "gpt-4o-mini-audio-preview-2024-12-17",
- "name": "GPT-4o-Mini Audio Preview 20241217",
+ "id": "gpt-4o-transcribe",
+ "name": "GPT-4o Transcribe",
"provider": "openai",
- "family": "gpt4o_mini_audio",
- "created_at": "2024-12-13 19:52:00 +0100",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "family": "gpt-4o-transcribe",
+ "created_at": null,
+ "context_window": 16000,
+ "max_output_tokens": 2000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "audio",
+ "text"
],
"output": [
- "text",
- "audio"
+ "embeddings",
+ "text"
]
},
- "capabilities": [
- "streaming",
- "speech_generation",
- "transcription"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -8131,34 +8198,39 @@
}
},
{
- "id": "gpt-4o-mini-realtime-preview",
- "name": "GPT-4o mini Realtime",
+ "id": "gpt-5",
+ "name": "GPT-5",
"provider": "openai",
- "family": "gpt-4o-mini-realtime-preview",
+ "family": "gpt-5",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 4096,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "image",
"text"
],
"output": [
- "audio",
"embeddings",
"text"
]
},
"capabilities": [
- "function_calling"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.6,
- "cached_input_per_million": 0.3,
- "output_per_million": 2.4
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.125,
+ "output_per_million": 10.0
+ },
+ "batch": {
+ "input_per_million": 0.625,
+ "output_per_million": 5.0
}
}
},
@@ -8168,34 +8240,39 @@
}
},
{
- "id": "gpt-4o-mini-realtime-preview-2024-12-17",
- "name": "GPT-4o mini Realtime",
+ "id": "gpt-5-2025-08-07",
+ "name": "GPT-5",
"provider": "openai",
- "family": "gpt-4o-mini-realtime-preview",
+ "family": "gpt-5",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 4096,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "image",
"text"
],
"output": [
- "audio",
"embeddings",
"text"
]
},
"capabilities": [
- "function_calling"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.6,
- "cached_input_per_million": 0.3,
- "output_per_million": 2.4
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.125,
+ "output_per_million": 10.0
+ },
+ "batch": {
+ "input_per_million": 0.625,
+ "output_per_million": 5.0
}
}
},
@@ -8205,16 +8282,17 @@
}
},
{
- "id": "gpt-4o-mini-search-preview",
- "name": "GPT-4o mini Search Preview",
+ "id": "gpt-5-chat-latest",
+ "name": "GPT-5 Chat",
"provider": "openai",
- "family": "gpt-4o-mini-search-preview",
+ "family": "gpt-5-chat-latest",
"created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -8222,14 +8300,13 @@
"text"
]
},
- "capabilities": [
- "structured_output"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.125,
+ "output_per_million": 10.0
}
}
},
@@ -8239,16 +8316,17 @@
}
},
{
- "id": "gpt-4o-mini-search-preview-2025-03-11",
- "name": "GPT-4o mini Search Preview",
+ "id": "gpt-5-mini",
+ "name": "GPT-5 mini",
"provider": "openai",
- "family": "gpt-4o-mini-search-preview",
+ "family": "gpt-5-mini",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -8257,13 +8335,20 @@
]
},
"capabilities": [
+ "batch",
+ "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6
+ "input_per_million": 0.25,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 2.0
+ },
+ "batch": {
+ "input_per_million": 0.125,
+ "output_per_million": 1.0
}
}
},
@@ -8273,17 +8358,17 @@
}
},
{
- "id": "gpt-4o-mini-transcribe",
- "name": "GPT-4o mini Transcribe",
- "provider": "openai",
- "family": "gpt-4o-mini-transcribe",
+ "id": "gpt-5-mini-2025-08-07",
+ "name": "GPT-5 mini",
+ "provider": "openai",
+ "family": "gpt-5-mini",
"created_at": null,
- "context_window": 16000,
- "max_output_tokens": 2000,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "image",
"text"
],
"output": [
@@ -8291,12 +8376,21 @@
"text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "batch",
+ "function_calling",
+ "structured_output"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 5.0
+ "input_per_million": 0.25,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 2.0
+ },
+ "batch": {
+ "input_per_million": 0.125,
+ "output_per_million": 1.0
}
}
},
@@ -8306,29 +8400,39 @@
}
},
{
- "id": "gpt-4o-mini-tts",
- "name": "GPT-4o mini TTS",
+ "id": "gpt-5-nano",
+ "name": "GPT-5 nano",
"provider": "openai",
- "family": "gpt-4o-mini-tts",
+ "family": "gpt-5-nano",
"created_at": null,
- "context_window": 2000,
- "max_output_tokens": null,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
- "audio",
- "embeddings"
+ "embeddings",
+ "text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "batch",
+ "function_calling",
+ "structured_output"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.6,
- "output_per_million": 12.0
+ "input_per_million": 0.05,
+ "cached_input_per_million": 0.005,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.025,
+ "output_per_million": 0.2
}
}
},
@@ -8338,34 +8442,39 @@
}
},
{
- "id": "gpt-4o-realtime-preview",
- "name": "GPT-4o Realtime",
+ "id": "gpt-5-nano-2025-08-07",
+ "name": "GPT-5 nano",
"provider": "openai",
- "family": "gpt-4o-realtime-preview",
+ "family": "gpt-5-nano",
"created_at": null,
- "context_window": 32000,
- "max_output_tokens": 4096,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "image",
"text"
],
"output": [
- "audio",
"embeddings",
"text"
]
},
"capabilities": [
- "function_calling"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 5.0,
- "cached_input_per_million": 2.5,
- "output_per_million": 20.0
+ "input_per_million": 0.05,
+ "cached_input_per_million": 0.005,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.025,
+ "output_per_million": 0.2
}
}
},
@@ -8375,30 +8484,31 @@
}
},
{
- "id": "gpt-4o-realtime-preview-2024-10-01",
- "name": "GPT-4o-Realtime Preview 20241001",
+ "id": "gpt-image-1",
+ "name": "GPT Image 1",
"provider": "openai",
- "family": "gpt4o_realtime",
- "created_at": "2024-09-24 00:49:26 +0200",
- "context_window": 128000,
- "max_output_tokens": 4096,
+ "family": "gpt-image-1",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
- "text"
+ "embeddings",
+ "image"
]
},
- "capabilities": [
- "streaming"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 5.0,
- "output_per_million": 20.0
+ "cached_input_per_million": 1.25,
+ "output_per_million": 40.0
}
}
},
@@ -8408,63 +8518,91 @@
}
},
{
- "id": "gpt-4o-realtime-preview-2024-12-17",
- "name": "GPT-4o-Realtime Preview 20241217",
+ "id": "gpt-oss-120b",
+ "name": "gpt-oss-120b",
"provider": "openai",
- "family": "gpt4o_realtime",
- "created_at": "2024-12-11 20:30:30 +0100",
- "context_window": 128000,
- "max_output_tokens": 4096,
+ "family": "gpt-oss-120b",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": 131072,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling",
+ "structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 5.0,
- "output_per_million": 20.0
- }
- }
+ "pricing": {},
+ "metadata": {}
+ },
+ {
+ "id": "gpt-oss-20b",
+ "name": "gpt-oss-20b",
+ "provider": "openai",
+ "family": "gpt-oss-20b",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": 131072,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "text"
+ ]
},
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
+ "capabilities": [
+ "batch",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {},
+ "metadata": {}
},
{
- "id": "gpt-4o-realtime-preview-2025-06-03",
- "name": "GPT-4o-Realtime Preview 20250603",
+ "id": "o1",
+ "name": "o1",
"provider": "openai",
- "family": "gpt4o_realtime",
- "created_at": "2025-06-03 01:43:58 +0200",
- "context_window": 128000,
- "max_output_tokens": 4096,
+ "family": "o1",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 5.0,
- "output_per_million": 20.0
+ "input_per_million": 15.0,
+ "cached_input_per_million": 7.5,
+ "output_per_million": 60.0
+ },
+ "batch": {
+ "input_per_million": 7.5,
+ "output_per_million": 30.0
}
}
},
@@ -8474,16 +8612,17 @@
}
},
{
- "id": "gpt-4o-search-preview",
- "name": "GPT-4o Search Preview",
+ "id": "o1-2024-12-17",
+ "name": "o1",
"provider": "openai",
- "family": "gpt-4o-search-preview",
+ "family": "o1",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -8492,13 +8631,20 @@
]
},
"capabilities": [
+ "batch",
+ "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 15.0,
+ "cached_input_per_million": 7.5,
+ "output_per_million": 60.0
+ },
+ "batch": {
+ "input_per_million": 7.5,
+ "output_per_million": 30.0
}
}
},
@@ -8508,13 +8654,13 @@
}
},
{
- "id": "gpt-4o-search-preview-2025-03-11",
- "name": "GPT-4o Search Preview",
+ "id": "o1-mini",
+ "name": "o1-mini",
"provider": "openai",
- "family": "gpt-4o-search-preview",
+ "family": "o1-mini",
"created_at": null,
"context_window": 128000,
- "max_output_tokens": 16384,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -8525,14 +8671,13 @@
"text"
]
},
- "capabilities": [
- "structured_output"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 1.1,
+ "cached_input_per_million": 0.55,
+ "output_per_million": 4.4
}
}
},
@@ -8542,17 +8687,16 @@
}
},
{
- "id": "gpt-4o-transcribe",
- "name": "GPT-4o Transcribe",
+ "id": "o1-mini-2024-09-12",
+ "name": "o1-mini",
"provider": "openai",
- "family": "gpt-4o-transcribe",
+ "family": "o1-mini",
"created_at": null,
- "context_window": 16000,
- "max_output_tokens": 2000,
+ "context_window": 128000,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
"text"
],
"output": [
@@ -8564,8 +8708,9 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 1.1,
+ "cached_input_per_million": 0.55,
+ "output_per_million": 4.4
}
}
},
@@ -8575,17 +8720,16 @@
}
},
{
- "id": "gpt-5",
- "name": "GPT-5",
+ "id": "o1-preview",
+ "name": "o1 Preview",
"provider": "openai",
- "family": "gpt-5",
+ "family": "o1-preview",
"created_at": null,
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "context_window": 128000,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -8594,40 +8738,31 @@
]
},
"capabilities": [
- "batch",
"function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.125,
- "output_per_million": 10.0
- },
- "batch": {
- "input_per_million": 0.625,
- "output_per_million": 5.0
+ "input_per_million": 15.0,
+ "cached_input_per_million": 7.5,
+ "output_per_million": 60.0
}
}
},
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
+ "metadata": {}
},
{
- "id": "gpt-5-2025-08-07",
- "name": "GPT-5",
+ "id": "o1-preview-2024-09-12",
+ "name": "o1 Preview",
"provider": "openai",
- "family": "gpt-5",
+ "family": "o1-preview",
"created_at": null,
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "context_window": 128000,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -8636,36 +8771,28 @@
]
},
"capabilities": [
- "batch",
"function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.125,
- "output_per_million": 10.0
- },
- "batch": {
- "input_per_million": 0.625,
- "output_per_million": 5.0
+ "input_per_million": 15.0,
+ "cached_input_per_million": 7.5,
+ "output_per_million": 60.0
}
}
},
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
+ "metadata": {}
},
{
- "id": "gpt-5-chat-latest",
- "name": "GPT-5 Chat",
+ "id": "o1-pro",
+ "name": "o1-pro",
"provider": "openai",
- "family": "gpt-5-chat-latest",
+ "family": "o1-pro",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -8677,13 +8804,20 @@
"text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "batch",
+ "function_calling",
+ "structured_output"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.125,
- "output_per_million": 10.0
+ "input_per_million": 150.0,
+ "output_per_million": 600.0
+ },
+ "batch": {
+ "input_per_million": 75.0,
+ "output_per_million": 300.0
}
}
},
@@ -8693,13 +8827,13 @@
}
},
{
- "id": "gpt-5-mini",
- "name": "GPT-5 mini",
+ "id": "o1-pro-2025-03-19",
+ "name": "o1-pro",
"provider": "openai",
- "family": "gpt-5-mini",
+ "family": "o1-pro",
"created_at": null,
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -8719,13 +8853,12 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "cached_input_per_million": 0.025,
- "output_per_million": 2.0
+ "input_per_million": 150.0,
+ "output_per_million": 600.0
},
"batch": {
- "input_per_million": 0.125,
- "output_per_million": 1.0
+ "input_per_million": 75.0,
+ "output_per_million": 300.0
}
}
},
@@ -8735,13 +8868,13 @@
}
},
{
- "id": "gpt-5-mini-2025-08-07",
- "name": "GPT-5 mini",
+ "id": "o3",
+ "name": "o3",
"provider": "openai",
- "family": "gpt-5-mini",
+ "family": "o3",
"created_at": null,
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -8761,13 +8894,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "cached_input_per_million": 0.025,
- "output_per_million": 2.0
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
},
"batch": {
- "input_per_million": 0.125,
- "output_per_million": 1.0
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
@@ -8777,13 +8910,13 @@
}
},
{
- "id": "gpt-5-nano",
- "name": "GPT-5 nano",
+ "id": "o3-2025-04-16",
+ "name": "o3",
"provider": "openai",
- "family": "gpt-5-nano",
+ "family": "o3",
"created_at": null,
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -8803,13 +8936,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.05,
- "cached_input_per_million": 0.005,
- "output_per_million": 0.4
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
},
"batch": {
- "input_per_million": 0.025,
- "output_per_million": 0.2
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
@@ -8819,13 +8952,13 @@
}
},
{
- "id": "gpt-5-nano-2025-08-07",
- "name": "GPT-5 nano",
+ "id": "o3-deep-research",
+ "name": "o3-deep-research",
"provider": "openai",
- "family": "gpt-5-nano",
+ "family": "o3-deep-research",
"created_at": null,
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -8838,20 +8971,18 @@
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.05,
- "cached_input_per_million": 0.005,
- "output_per_million": 0.4
+ "input_per_million": 10.0,
+ "cached_input_per_million": 2.5,
+ "output_per_million": 40.0
},
"batch": {
- "input_per_million": 0.025,
- "output_per_million": 0.2
+ "input_per_million": 5.0,
+ "output_per_million": 20.0
}
}
},
@@ -8861,13 +8992,13 @@
}
},
{
- "id": "gpt-image-1",
- "name": "GPT Image 1",
+ "id": "o3-deep-research-2025-06-26",
+ "name": "o3-deep-research",
"provider": "openai",
- "family": "gpt-image-1",
+ "family": "o3-deep-research",
"created_at": null,
- "context_window": null,
- "max_output_tokens": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -8876,16 +9007,22 @@
],
"output": [
"embeddings",
- "image"
+ "text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "batch"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 5.0,
- "cached_input_per_million": 1.25,
+ "input_per_million": 10.0,
+ "cached_input_per_million": 2.5,
"output_per_million": 40.0
+ },
+ "batch": {
+ "input_per_million": 5.0,
+ "output_per_million": 20.0
}
}
},
@@ -8895,13 +9032,13 @@
}
},
{
- "id": "gpt-oss-120b",
- "name": "gpt-oss-120b",
+ "id": "o3-mini",
+ "name": "o3-mini",
"provider": "openai",
- "family": "gpt-oss-120b",
+ "family": "o3-mini",
"created_at": null,
- "context_window": 131072,
- "max_output_tokens": 131072,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -8917,17 +9054,32 @@
"function_calling",
"structured_output"
],
- "pricing": {},
- "metadata": {}
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.1,
+ "cached_input_per_million": 0.55,
+ "output_per_million": 4.4
+ },
+ "batch": {
+ "input_per_million": 0.55,
+ "output_per_million": 2.2
+ }
+ }
+ },
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
},
{
- "id": "gpt-oss-20b",
- "name": "gpt-oss-20b",
+ "id": "o3-mini-2025-01-31",
+ "name": "o3-mini",
"provider": "openai",
- "family": "gpt-oss-20b",
+ "family": "o3-mini",
"created_at": null,
- "context_window": 131072,
- "max_output_tokens": 131072,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -8943,14 +9095,29 @@
"function_calling",
"structured_output"
],
- "pricing": {},
- "metadata": {}
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.1,
+ "cached_input_per_million": 0.55,
+ "output_per_million": 4.4
+ },
+ "batch": {
+ "input_per_million": 0.55,
+ "output_per_million": 2.2
+ }
+ }
+ },
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
},
{
- "id": "o1",
- "name": "o1",
+ "id": "o3-pro",
+ "name": "o3-pro",
"provider": "openai",
- "family": "o1",
+ "family": "o3-pro",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 100000,
@@ -8973,13 +9140,12 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 7.5,
- "output_per_million": 60.0
+ "input_per_million": 20.0,
+ "output_per_million": 80.0
},
"batch": {
- "input_per_million": 7.5,
- "output_per_million": 30.0
+ "input_per_million": 10.0,
+ "output_per_million": 40.0
}
}
},
@@ -8989,10 +9155,10 @@
}
},
{
- "id": "o1-2024-12-17",
- "name": "o1",
+ "id": "o3-pro-2025-06-10",
+ "name": "o3-pro",
"provider": "openai",
- "family": "o1",
+ "family": "o3-pro",
"created_at": null,
"context_window": 200000,
"max_output_tokens": 100000,
@@ -9015,13 +9181,12 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 7.5,
- "output_per_million": 60.0
+ "input_per_million": 20.0,
+ "output_per_million": 80.0
},
"batch": {
- "input_per_million": 7.5,
- "output_per_million": 30.0
+ "input_per_million": 10.0,
+ "output_per_million": 40.0
}
}
},
@@ -9031,16 +9196,17 @@
}
},
{
- "id": "o1-mini",
- "name": "o1-mini",
+ "id": "o4-mini",
+ "name": "o4-mini",
"provider": "openai",
- "family": "o1-mini",
+ "family": "o4-mini",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 65536,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -9048,13 +9214,21 @@
"text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "batch",
+ "function_calling",
+ "structured_output"
+ ],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 1.1,
- "cached_input_per_million": 0.55,
+ "cached_input_per_million": 0.275,
"output_per_million": 4.4
+ },
+ "batch": {
+ "input_per_million": 0.55,
+ "output_per_million": 2.2
}
}
},
@@ -9064,16 +9238,17 @@
}
},
{
- "id": "o1-mini-2024-09-12",
- "name": "o1-mini",
+ "id": "o4-mini-2025-04-16",
+ "name": "o4-mini",
"provider": "openai",
- "family": "o1-mini",
+ "family": "o4-mini",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 65536,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -9081,13 +9256,21 @@
"text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "batch",
+ "function_calling",
+ "structured_output"
+ ],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 1.1,
- "cached_input_per_million": 0.55,
+ "cached_input_per_million": 0.275,
"output_per_million": 4.4
+ },
+ "batch": {
+ "input_per_million": 0.55,
+ "output_per_million": 2.2
}
}
},
@@ -9097,16 +9280,17 @@
}
},
{
- "id": "o1-preview",
- "name": "o1 Preview",
+ "id": "o4-mini-deep-research",
+ "name": "o4-mini-deep-research",
"provider": "openai",
- "family": "o1-preview",
+ "family": "o4-mini-deep-research",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 32768,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -9115,31 +9299,38 @@
]
},
"capabilities": [
- "function_calling",
- "structured_output"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 7.5,
- "output_per_million": 60.0
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
+ },
+ "batch": {
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
- "metadata": {}
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
},
{
- "id": "o1-preview-2024-09-12",
- "name": "o1 Preview",
+ "id": "o4-mini-deep-research-2025-06-26",
+ "name": "o4-mini-deep-research",
"provider": "openai",
- "family": "o1-preview",
+ "family": "o4-mini-deep-research",
"created_at": null,
- "context_window": 128000,
- "max_output_tokens": 32768,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -9148,69 +9339,61 @@
]
},
"capabilities": [
- "function_calling",
- "structured_output"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 7.5,
- "output_per_million": 60.0
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
+ },
+ "batch": {
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
- "metadata": {}
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
},
{
- "id": "o1-pro",
- "name": "o1-pro",
+ "id": "omni-moderation-2024-09-26",
+ "name": "Omni Moderation 20240926",
"provider": "openai",
- "family": "o1-pro",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "moderation",
+ "created_at": "2024-11-28 06:07:46 +1100",
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
- "embeddings",
- "text"
+ "text",
+ "moderation"
]
},
- "capabilities": [
- "batch",
- "function_calling",
- "structured_output"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 150.0,
- "output_per_million": 600.0
- },
- "batch": {
- "input_per_million": 75.0,
- "output_per_million": 300.0
- }
- }
- },
+ "capabilities": [],
+ "pricing": {},
"metadata": {
"object": "model",
"owned_by": "system"
}
},
{
- "id": "o1-pro-2025-03-19",
- "name": "o1-pro",
+ "id": "omni-moderation-latest",
+ "name": "omni-moderation",
"provider": "openai",
- "family": "o1-pro",
+ "family": "omni-moderation-latest",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -9222,40 +9405,24 @@
"text"
]
},
- "capabilities": [
- "batch",
- "function_calling",
- "structured_output"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 150.0,
- "output_per_million": 600.0
- },
- "batch": {
- "input_per_million": 75.0,
- "output_per_million": 300.0
- }
- }
- },
+ "capabilities": [],
+ "pricing": {},
"metadata": {
"object": "model",
"owned_by": "system"
}
},
{
- "id": "o3",
- "name": "o3",
+ "id": "text-embedding-3-large",
+ "name": "text-embedding-3-large",
"provider": "openai",
- "family": "o3",
+ "family": "text-embedding-3-large",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -9264,20 +9431,23 @@
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "cached_input_per_million": 0.5,
- "output_per_million": 8.0
+ "input_per_million": 0.13
},
"batch": {
- "input_per_million": 1.0,
- "output_per_million": 4.0
+ "input_per_million": 0.065
+ }
+ },
+ "embeddings": {
+ "standard": {
+ "input_per_million": 0.13
+ },
+ "batch": {
+ "input_per_million": 0.065
}
}
},
@@ -9287,17 +9457,16 @@
}
},
{
- "id": "o3-2025-04-16",
- "name": "o3",
+ "id": "text-embedding-3-small",
+ "name": "text-embedding-3-small",
"provider": "openai",
- "family": "o3",
+ "family": "text-embedding-3-small",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -9306,20 +9475,23 @@
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "cached_input_per_million": 0.5,
- "output_per_million": 8.0
+ "input_per_million": 0.02
},
"batch": {
- "input_per_million": 1.0,
- "output_per_million": 4.0
+ "input_per_million": 0.01
+ }
+ },
+ "embeddings": {
+ "standard": {
+ "input_per_million": 0.02
+ },
+ "batch": {
+ "input_per_million": 0.01
}
}
},
@@ -9329,17 +9501,16 @@
}
},
{
- "id": "o3-deep-research",
- "name": "o3-deep-research",
+ "id": "text-embedding-ada-002",
+ "name": "text-embedding-ada-002",
"provider": "openai",
- "family": "o3-deep-research",
+ "family": "text-embedding-ada-002",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -9353,33 +9524,37 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 10.0,
- "cached_input_per_million": 2.5,
- "output_per_million": 40.0
+ "input_per_million": 0.1
},
"batch": {
- "input_per_million": 5.0,
- "output_per_million": 20.0
+ "input_per_million": 0.05
}
- }
- },
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
+ },
+ "embeddings": {
+ "standard": {
+ "input_per_million": 0.1
+ },
+ "batch": {
+ "input_per_million": 0.05
+ }
+ }
+ },
+ "metadata": {
+ "object": "model",
+ "owned_by": "openai-internal"
+ }
},
{
- "id": "o3-deep-research-2025-06-26",
- "name": "o3-deep-research",
+ "id": "text-moderation-latest",
+ "name": "text-moderation",
"provider": "openai",
- "family": "o3-deep-research",
+ "family": "text-moderation-latest",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "context_window": null,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -9387,35 +9562,18 @@
"text"
]
},
- "capabilities": [
- "batch"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 10.0,
- "cached_input_per_million": 2.5,
- "output_per_million": 40.0
- },
- "batch": {
- "input_per_million": 5.0,
- "output_per_million": 20.0
- }
- }
- },
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
+ "capabilities": [],
+ "pricing": {},
+ "metadata": {}
},
{
- "id": "o3-mini",
- "name": "o3-mini",
+ "id": "text-moderation-stable",
+ "name": "text-moderation-stable",
"provider": "openai",
- "family": "o3-mini",
+ "family": "text-moderation-stable",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "context_window": null,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -9426,103 +9584,68 @@
"text"
]
},
- "capabilities": [
- "batch",
- "function_calling",
- "structured_output"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 1.1,
- "cached_input_per_million": 0.55,
- "output_per_million": 4.4
- },
- "batch": {
- "input_per_million": 0.55,
- "output_per_million": 2.2
- }
- }
- },
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
+ "capabilities": [],
+ "pricing": {},
+ "metadata": {}
},
{
- "id": "o3-mini-2025-01-31",
- "name": "o3-mini",
+ "id": "tts-1",
+ "name": "TTS-1",
"provider": "openai",
- "family": "o3-mini",
+ "family": "tts-1",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "embeddings",
- "text"
+ "audio",
+ "embeddings"
]
},
- "capabilities": [
- "batch",
- "function_calling",
- "structured_output"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.1,
- "cached_input_per_million": 0.55,
- "output_per_million": 4.4
- },
- "batch": {
- "input_per_million": 0.55,
- "output_per_million": 2.2
+ "input_per_million": 15.0
}
}
},
"metadata": {
"object": "model",
- "owned_by": "system"
+ "owned_by": "openai-internal"
}
},
{
- "id": "o3-pro",
- "name": "o3-pro",
+ "id": "tts-1-1106",
+ "name": "TTS-1 1106",
"provider": "openai",
- "family": "o3-pro",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "tts1",
+ "created_at": "2023-11-04 10:14:01 +1100",
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
- "text"
+ "text",
+ "audio"
],
"output": [
- "embeddings",
- "text"
+ "text",
+ "audio"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 20.0,
- "output_per_million": 80.0
- },
- "batch": {
- "input_per_million": 10.0,
- "output_per_million": 40.0
+ "input_per_million": 15.0,
+ "output_per_million": 15.0
}
}
},
@@ -9532,38 +9655,28 @@
}
},
{
- "id": "o3-pro-2025-06-10",
- "name": "o3-pro",
+ "id": "tts-1-hd",
+ "name": "TTS-1 HD",
"provider": "openai",
- "family": "o3-pro",
+ "family": "tts-1-hd",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
- "embeddings",
- "text"
+ "audio",
+ "embeddings"
]
},
- "capabilities": [
- "batch",
- "function_calling",
- "structured_output"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 20.0,
- "output_per_million": 80.0
- },
- "batch": {
- "input_per_million": 10.0,
- "output_per_million": 40.0
+ "output_per_million": 30.0
}
}
},
@@ -9573,39 +9686,32 @@
}
},
{
- "id": "o4-mini",
- "name": "o4-mini",
+ "id": "tts-1-hd-1106",
+ "name": "TTS-1 HD 1106",
"provider": "openai",
- "family": "o4-mini",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "tts1_hd",
+ "created_at": "2023-11-04 10:18:53 +1100",
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
- "text"
+ "text",
+ "audio"
],
"output": [
- "embeddings",
- "text"
+ "text",
+ "audio"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.1,
- "cached_input_per_million": 0.275,
- "output_per_million": 4.4
- },
- "batch": {
- "input_per_million": 0.55,
- "output_per_million": 2.2
+ "input_per_million": 30.0,
+ "output_per_million": 30.0
}
}
},
@@ -9615,319 +9721,347 @@
}
},
{
- "id": "o4-mini-2025-04-16",
- "name": "o4-mini",
+ "id": "whisper-1",
+ "name": "Whisper",
"provider": "openai",
- "family": "o4-mini",
+ "family": "whisper-1",
"created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
- "text"
+ "audio"
],
"output": [
+ "audio",
"embeddings",
"text"
]
},
- "capabilities": [
- "batch",
- "function_calling",
- "structured_output"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.1,
- "cached_input_per_million": 0.275,
- "output_per_million": 4.4
- },
- "batch": {
- "input_per_million": 0.55,
- "output_per_million": 2.2
+ "input_per_million": 0.006
}
}
},
"metadata": {
"object": "model",
- "owned_by": "system"
+ "owned_by": "openai-internal"
}
},
{
- "id": "o4-mini-deep-research",
- "name": "o4-mini-deep-research",
- "provider": "openai",
- "family": "o4-mini-deep-research",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "id": "agentica-org/deepcoder-14b-preview",
+ "name": "Agentica: Deepcoder 14B Preview",
+ "provider": "openrouter",
+ "family": "agentica-org",
+ "created_at": "2025-04-14 00:43:15 +1000",
+ "context_window": 96000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
- "embeddings",
"text"
]
},
"capabilities": [
- "batch"
+ "streaming",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "cached_input_per_million": 0.5,
- "output_per_million": 8.0
- },
- "batch": {
- "input_per_million": 1.0,
- "output_per_million": 4.0
+ "input_per_million": 0.015,
+ "output_per_million": 0.015
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "system"
+ "description": "DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Other",
+ "instruct_type": "deepseek-r1"
+ },
+ "top_provider": {
+ "context_length": 96000,
+ "max_completion_tokens": null,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
+ "max_tokens",
+ "min_p",
+ "presence_penalty",
+ "reasoning",
+ "repetition_penalty",
+ "seed",
+ "stop",
+ "temperature",
+ "top_k",
+ "top_logprobs",
+ "top_p"
+ ]
}
},
{
- "id": "o4-mini-deep-research-2025-06-26",
- "name": "o4-mini-deep-research",
- "provider": "openai",
- "family": "o4-mini-deep-research",
- "created_at": null,
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "id": "agentica-org/deepcoder-14b-preview:free",
+ "name": "Agentica: Deepcoder 14B Preview (free)",
+ "provider": "openrouter",
+ "family": "agentica-org",
+ "created_at": "2025-04-14 00:43:15 +1000",
+ "context_window": 96000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
- "embeddings",
"text"
]
},
"capabilities": [
- "batch"
+ "streaming",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 2.0,
- "cached_input_per_million": 0.5,
- "output_per_million": 8.0
- },
- "batch": {
- "input_per_million": 1.0,
- "output_per_million": 4.0
- }
- }
- },
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
- },
- {
- "id": "omni-moderation-2024-09-26",
- "name": "Omni Moderation 20240926",
- "provider": "openai",
- "family": "moderation",
- "created_at": "2024-11-27 20:07:46 +0100",
- "context_window": null,
- "max_output_tokens": null,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "text",
- "image",
- "pdf"
- ],
- "output": [
- "text",
- "moderation"
- ]
- },
- "capabilities": [],
"pricing": {},
"metadata": {
- "object": "model",
- "owned_by": "system"
- }
- },
- {
- "id": "omni-moderation-latest",
- "name": "omni-moderation",
- "provider": "openai",
- "family": "omni-moderation-latest",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "image",
- "text"
- ],
- "output": [
- "embeddings",
- "text"
+ "description": "DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Other",
+ "instruct_type": "deepseek-r1"
+ },
+ "top_provider": {
+ "context_length": 96000,
+ "max_completion_tokens": null,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
+ "max_tokens",
+ "min_p",
+ "presence_penalty",
+ "reasoning",
+ "repetition_penalty",
+ "seed",
+ "stop",
+ "temperature",
+ "top_k",
+ "top_logprobs",
+ "top_p"
]
- },
- "capabilities": [],
- "pricing": {},
- "metadata": {
- "object": "model",
- "owned_by": "system"
}
},
{
- "id": "text-embedding-3-large",
- "name": "text-embedding-3-large",
- "provider": "openai",
- "family": "text-embedding-3-large",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
+ "id": "ai21/jamba-large-1.7",
+ "name": "AI21: Jamba Large 1.7",
+ "provider": "openrouter",
+ "family": "ai21",
+ "created_at": "2025-08-09 02:03:40 +1000",
+ "context_window": 256000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "embeddings",
"text"
]
},
"capabilities": [
- "batch"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.13
- },
- "batch": {
- "input_per_million": 0.065
- }
- },
- "embeddings": {
- "standard": {
- "input_per_million": 0.13
- },
- "batch": {
- "input_per_million": 0.065
+ "input_per_million": 2.0,
+ "output_per_million": 8.0
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "system"
+ "description": "Jamba Large 1.7 is the latest model in the Jamba open family, offering improvements in grounding, instruction-following, and overall efficiency. Built on a hybrid SSM-Transformer architecture with a 256K context window, it delivers more accurate, contextually grounded responses and better steerability than previous versions.",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Other",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 256000,
+ "max_completion_tokens": 4096,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "max_tokens",
+ "response_format",
+ "stop",
+ "temperature",
+ "tool_choice",
+ "tools",
+ "top_p"
+ ]
}
},
{
- "id": "text-embedding-3-small",
- "name": "text-embedding-3-small",
- "provider": "openai",
- "family": "text-embedding-3-small",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
+ "id": "ai21/jamba-mini-1.7",
+ "name": "AI21: Jamba Mini 1.7",
+ "provider": "openrouter",
+ "family": "ai21",
+ "created_at": "2025-08-09 02:30:01 +1000",
+ "context_window": 256000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "embeddings",
"text"
]
},
"capabilities": [
- "batch"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.02
- },
- "batch": {
- "input_per_million": 0.01
- }
- },
- "embeddings": {
- "standard": {
- "input_per_million": 0.02
- },
- "batch": {
- "input_per_million": 0.01
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.39999999999999997
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "system"
+ "description": "Jamba Mini 1.7 is a compact and efficient member of the Jamba open model family, incorporating key improvements in grounding and instruction-following while maintaining the benefits of the SSM-Transformer hybrid architecture and 256K context window. Despite its compact size, it delivers accurate, contextually grounded responses and improved steerability.",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Other",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 256000,
+ "max_completion_tokens": 4096,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "max_tokens",
+ "response_format",
+ "stop",
+ "temperature",
+ "tool_choice",
+ "tools",
+ "top_p"
+ ]
}
},
{
- "id": "text-embedding-ada-002",
- "name": "text-embedding-ada-002",
- "provider": "openai",
- "family": "text-embedding-ada-002",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
+ "id": "aion-labs/aion-1.0",
+ "name": "AionLabs: Aion-1.0",
+ "provider": "openrouter",
+ "family": "aion-labs",
+ "created_at": "2025-02-05 06:32:37 +1100",
+ "context_window": 131072,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "embeddings",
"text"
]
},
"capabilities": [
- "batch"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1
- },
- "batch": {
- "input_per_million": 0.05
- }
- },
- "embeddings": {
- "standard": {
- "input_per_million": 0.1
- },
- "batch": {
- "input_per_million": 0.05
+ "input_per_million": 4.0,
+ "output_per_million": 8.0
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "openai-internal"
+ "description": "Aion-1.0 is a multi-model system designed for high performance across various tasks, including reasoning and coding. It is built on DeepSeek-R1, augmented with additional models and techniques such as Tree of Thoughts (ToT) and Mixture of Experts (MoE). It is Aion Lab's most powerful reasoning model.",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Other",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 131072,
+ "max_completion_tokens": 32768,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "include_reasoning",
+ "max_tokens",
+ "reasoning",
+ "temperature",
+ "top_p"
+ ]
}
},
{
- "id": "text-moderation-latest",
- "name": "text-moderation",
- "provider": "openai",
- "family": "text-moderation-latest",
- "created_at": null,
- "context_window": null,
+ "id": "aion-labs/aion-1.0-mini",
+ "name": "AionLabs: Aion-1.0-Mini",
+ "provider": "openrouter",
+ "family": "aion-labs",
+ "created_at": "2025-02-05 06:25:07 +1100",
+ "context_window": 131072,
"max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
@@ -9935,21 +10069,55 @@
"text"
],
"output": [
- "embeddings",
"text"
]
},
- "capabilities": [],
- "pricing": {},
- "metadata": {}
+ "capabilities": [
+ "streaming"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.7,
+ "output_per_million": 1.4
+ }
+ }
+ },
+ "metadata": {
+ "description": "Aion-1.0-Mini 32B parameter model is a distilled version of the DeepSeek-R1 model, designed for strong performance in reasoning domains such as mathematics, coding, and logic. It is a modified variant of a FuseAI model that outperforms R1-Distill-Qwen-32B and R1-Distill-Llama-70B, with benchmark results available on its [Hugging Face page](https://huggingface.co/FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview), independently replicated for verification.",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Other",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 131072,
+ "max_completion_tokens": 32768,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "include_reasoning",
+ "max_tokens",
+ "reasoning",
+ "temperature",
+ "top_p"
+ ]
+ }
},
{
- "id": "text-moderation-stable",
- "name": "text-moderation-stable",
- "provider": "openai",
- "family": "text-moderation-stable",
- "created_at": null,
- "context_window": null,
+ "id": "aion-labs/aion-rp-llama-3.1-8b",
+ "name": "AionLabs: Aion-RP 1.0 (8B)",
+ "provider": "openrouter",
+ "family": "aion-labs",
+ "created_at": "2025-02-05 06:18:38 +1100",
+ "context_window": 32768,
"max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
@@ -9957,62 +10125,61 @@
"text"
],
"output": [
- "embeddings",
- "text"
- ]
- },
- "capabilities": [],
- "pricing": {},
- "metadata": {}
- },
- {
- "id": "tts-1",
- "name": "TTS-1",
- "provider": "openai",
- "family": "tts-1",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
"text"
- ],
- "output": [
- "audio",
- "embeddings"
]
},
- "capabilities": [],
+ "capabilities": [
+ "streaming"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.19999999999999998
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "openai-internal"
+ "description": "Aion-RP-Llama-3.1-8B ranks the highest in the character evaluation portion of the RPBench-Auto benchmark, a roleplaying-specific variant of Arena-Hard-Auto, where LLMs evaluate each other’s responses. It is a fine-tuned base model rather than an instruct model, designed to produce more natural and varied writing.",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Other",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 32768,
+ "max_completion_tokens": 32768,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "max_tokens",
+ "temperature",
+ "top_p"
+ ]
}
},
{
- "id": "tts-1-1106",
- "name": "TTS-1 1106",
- "provider": "openai",
- "family": "tts1",
- "created_at": "2023-11-04 00:14:01 +0100",
- "context_window": null,
- "max_output_tokens": null,
+ "id": "alfredpros/codellama-7b-instruct-solidity",
+ "name": "AlfredPros: CodeLLaMa 7B Instruct Solidity",
+ "provider": "openrouter",
+ "family": "alfredpros",
+ "created_at": "2025-04-15 00:44:34 +1000",
+ "context_window": 8192,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "text"
],
"output": [
- "text",
- "audio"
+ "text"
]
},
"capabilities": [
@@ -10021,126 +10188,242 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 15.0
+ "input_per_million": 0.7,
+ "output_per_million": 1.1
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "system"
+ "description": "A finetuned 7 billion parameters Code LLaMA - Instruct model to generate Solidity smart contract using 4-bit QLoRA finetuning provided by PEFT library.",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Other",
+ "instruct_type": "alpaca"
+ },
+ "top_provider": {
+ "context_length": 8192,
+ "max_completion_tokens": 8192,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "frequency_penalty",
+ "max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "seed",
+ "stop",
+ "temperature",
+ "top_k",
+ "top_p"
+ ]
}
},
{
- "id": "tts-1-hd",
- "name": "TTS-1 HD",
- "provider": "openai",
- "family": "tts-1-hd",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
+ "id": "alpindale/goliath-120b",
+ "name": "Goliath 120B",
+ "provider": "openrouter",
+ "family": "alpindale",
+ "created_at": "2023-11-10 11:00:00 +1100",
+ "context_window": 6144,
+ "max_output_tokens": 512,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "audio",
- "embeddings"
+ "text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "streaming",
+ "structured_output",
+ "predicted_outputs"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 30.0
+ "input_per_million": 4.0,
+ "output_per_million": 5.5
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "system"
+ "description": "A large LLM created by combining two fine-tuned Llama 70B models into one 120B model. Combines Xwin and Euryale.\n\nCredits to\n- [@chargoddard](https://huggingface.co/chargoddard) for developing the framework used to merge the model - [mergekit](https://github.com/cg123/mergekit).\n- [@Undi95](https://huggingface.co/Undi95) for helping with the merge ratios.\n\n#merge",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Llama2",
+ "instruct_type": "airoboros"
+ },
+ "top_provider": {
+ "context_length": 6144,
+ "max_completion_tokens": 512,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
+ "max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
+ "temperature",
+ "top_a",
+ "top_k",
+ "top_p"
+ ]
}
},
{
- "id": "tts-1-hd-1106",
- "name": "TTS-1 HD 1106",
- "provider": "openai",
- "family": "tts1_hd",
- "created_at": "2023-11-04 00:18:53 +0100",
- "context_window": null,
- "max_output_tokens": null,
+ "id": "amazon/nova-lite-v1",
+ "name": "Amazon: Nova Lite 1.0",
+ "provider": "openrouter",
+ "family": "amazon",
+ "created_at": "2024-12-06 09:22:43 +1100",
+ "context_window": 300000,
+ "max_output_tokens": 5120,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text",
- "audio"
+ "image"
],
"output": [
- "text",
- "audio"
+ "text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 30.0,
- "output_per_million": 30.0
+ "input_per_million": 0.06,
+ "output_per_million": 0.24
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "system"
+ "description": "Amazon Nova Lite 1.0 is a very low-cost multimodal model from Amazon that focused on fast processing of image, video, and text inputs to generate text output. Amazon Nova Lite can handle real-time customer interactions, document analysis, and visual question-answering tasks with high accuracy.\n\nWith an input context of 300K tokens, it can analyze multiple images or up to 30 minutes of video in a single input.",
+ "architecture": {
+ "modality": "text+image->text",
+ "input_modalities": [
+ "text",
+ "image"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Nova",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 300000,
+ "max_completion_tokens": 5120,
+ "is_moderated": true
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "max_tokens",
+ "stop",
+ "temperature",
+ "tools",
+ "top_k",
+ "top_p"
+ ]
}
},
{
- "id": "whisper-1",
- "name": "Whisper",
- "provider": "openai",
- "family": "whisper-1",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
+ "id": "amazon/nova-micro-v1",
+ "name": "Amazon: Nova Micro 1.0",
+ "provider": "openrouter",
+ "family": "amazon",
+ "created_at": "2024-12-06 09:20:37 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 5120,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio"
+ "text"
],
"output": [
- "audio",
- "embeddings",
"text"
]
},
- "capabilities": [],
+ "capabilities": [
+ "streaming",
+ "function_calling"
+ ],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.006
+ "input_per_million": 0.035,
+ "output_per_million": 0.14
}
}
},
"metadata": {
- "object": "model",
- "owned_by": "openai-internal"
+ "description": "Amazon Nova Micro 1.0 is a text-only model that delivers the lowest latency responses in the Amazon Nova family of models at a very low cost. With a context length of 128K tokens and optimized for speed and cost, Amazon Nova Micro excels at tasks such as text summarization, translation, content classification, interactive chat, and brainstorming. It has simple mathematical reasoning and coding abilities.",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Nova",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 128000,
+ "max_completion_tokens": 5120,
+ "is_moderated": true
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "max_tokens",
+ "stop",
+ "temperature",
+ "tools",
+ "top_k",
+ "top_p"
+ ]
}
},
{
- "id": "agentica-org/deepcoder-14b-preview",
- "name": "Agentica: Deepcoder 14B Preview",
+ "id": "amazon/nova-pro-v1",
+ "name": "Amazon: Nova Pro 1.0",
"provider": "openrouter",
- "family": "agentica-org",
- "created_at": "2025-04-13 16:43:15 +0200",
- "context_window": 96000,
- "max_output_tokens": null,
+ "family": "amazon",
+ "created_at": "2024-12-06 09:05:03 +1100",
+ "context_window": 300000,
+ "max_output_tokens": 5120,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -10148,61 +10431,53 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.015,
- "output_per_million": 0.015
+ "input_per_million": 0.7999999999999999,
+ "output_per_million": 3.1999999999999997
}
}
},
"metadata": {
- "description": "DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini",
+ "description": "Amazon Nova Pro 1.0 is a capable multimodal model from Amazon focused on providing a combination of accuracy, speed, and cost for a wide range of tasks. As of December 2024, it achieves state-of-the-art performance on key benchmarks including visual question answering (TextVQA) and video understanding (VATEX).\n\nAmazon Nova Pro demonstrates strong capabilities in processing both visual and textual information and at analyzing financial documents.\n\n**NOTE**: Video input is not supported at this time.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Nova",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 96000,
- "max_completion_tokens": null,
- "is_moderated": false
+ "context_length": 300000,
+ "max_completion_tokens": 5120,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
- "seed",
"stop",
"temperature",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "agentica-org/deepcoder-14b-preview:free",
- "name": "Agentica: Deepcoder 14B Preview (free)",
+ "id": "anthracite-org/magnum-v2-72b",
+ "name": "Magnum v2 72B",
"provider": "openrouter",
- "family": "agentica-org",
- "created_at": "2025-04-13 16:43:15 +0200",
- "context_window": 96000,
+ "family": "anthracite-org",
+ "created_at": "2024-09-30 10:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -10217,9 +10492,16 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 3.0
+ }
+ }
+ },
"metadata": {
- "description": "DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini",
+ "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the seventh in a family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -10228,42 +10510,38 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Qwen",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 96000,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
"seed",
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "ai21/jamba-large-1.7",
- "name": "AI21: Jamba Large 1.7",
+ "id": "anthracite-org/magnum-v4-72b",
+ "name": "Magnum v4 72B",
"provider": "openrouter",
- "family": "ai21",
- "created_at": "2025-08-08 18:03:40 +0200",
- "context_window": 256000,
- "max_output_tokens": 4096,
+ "family": "anthracite-org",
+ "created_at": "2024-10-22 11:00:00 +1100",
+ "context_window": 16384,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -10275,19 +10553,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 2.0,
- "output_per_million": 8.0
+ "output_per_million": 5.0
}
}
},
"metadata": {
- "description": "Jamba Large 1.7 is the latest model in the Jamba open family, offering improvements in grounding, instruction-following, and overall efficiency. Built on a hybrid SSM-Transformer architecture with a 256K context window, it delivers more accurate, contextually grounded responses and better steerability than previous versions.",
+ "description": "This is a series of models designed to replicate the prose quality of the Claude 3 models, specifically Sonnet(https://openrouter.ai/anthropic/claude-3.5-sonnet) and Opus(https://openrouter.ai/anthropic/claude-3-opus).\n\nThe model is fine-tuned on top of [Qwen2.5 72B](https://openrouter.ai/qwen/qwen-2.5-72b-instruct).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -10296,38 +10573,44 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 256000,
- "max_completion_tokens": 4096,
+ "context_length": 16384,
+ "max_completion_tokens": 2048,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
- "response_format",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
+ "top_a",
+ "top_k",
"top_p"
]
}
},
{
- "id": "ai21/jamba-mini-1.7",
- "name": "AI21: Jamba Mini 1.7",
+ "id": "anthropic/claude-3-haiku",
+ "name": "Anthropic: Claude 3 Haiku",
"provider": "openrouter",
- "family": "ai21",
- "created_at": "2025-08-08 18:30:01 +0200",
- "context_window": 256000,
+ "family": "anthropic",
+ "created_at": "2024-03-13 11:00:00 +1100",
+ "context_window": 200000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -10335,286 +10618,314 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.39999999999999997
+ "input_per_million": 0.25,
+ "output_per_million": 1.25,
+ "cached_input_per_million": 0.03
}
}
},
"metadata": {
- "description": "Jamba Mini 1.7 is a compact and efficient member of the Jamba open model family, incorporating key improvements in grounding and instruction-following while maintaining the benefits of the SSM-Transformer hybrid architecture and 256K context window. Despite its compact size, it delivers accurate, contextually grounded responses and improved steerability.",
+ "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Claude",
"instruct_type": null
},
"top_provider": {
- "context_length": 256000,
+ "context_length": 200000,
"max_completion_tokens": 4096,
- "is_moderated": false
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"max_tokens",
- "response_format",
"stop",
"temperature",
"tool_choice",
"tools",
+ "top_k",
"top_p"
]
}
},
{
- "id": "aion-labs/aion-1.0",
- "name": "AionLabs: Aion-1.0",
+ "id": "anthropic/claude-3-opus",
+ "name": "Anthropic: Claude 3 Opus",
"provider": "openrouter",
- "family": "aion-labs",
- "created_at": "2025-02-04 20:32:37 +0100",
- "context_window": 131072,
- "max_output_tokens": 32768,
+ "family": "anthropic",
+ "created_at": "2024-03-05 11:00:00 +1100",
+ "context_window": 200000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 4.0,
- "output_per_million": 8.0
+ "input_per_million": 15.0,
+ "output_per_million": 75.0,
+ "cached_input_per_million": 1.5
}
}
},
"metadata": {
- "description": "Aion-1.0 is a multi-model system designed for high performance across various tasks, including reasoning and coding. It is built on DeepSeek-R1, augmented with additional models and techniques such as Tree of Thoughts (ToT) and Mixture of Experts (MoE). It is Aion Lab's most powerful reasoning model.",
+ "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Claude",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 32768,
- "is_moderated": false
+ "context_length": 200000,
+ "max_completion_tokens": 4096,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
"max_tokens",
- "reasoning",
+ "stop",
"temperature",
+ "tool_choice",
+ "tools",
+ "top_k",
"top_p"
]
}
},
{
- "id": "aion-labs/aion-1.0-mini",
- "name": "AionLabs: Aion-1.0-Mini",
+ "id": "anthropic/claude-3.5-haiku",
+ "name": "Anthropic: Claude 3.5 Haiku",
"provider": "openrouter",
- "family": "aion-labs",
- "created_at": "2025-02-04 20:25:07 +0100",
- "context_window": 131072,
- "max_output_tokens": 32768,
+ "family": "anthropic",
+ "created_at": "2024-11-04 11:00:00 +1100",
+ "context_window": 200000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7,
- "output_per_million": 1.4
+ "input_per_million": 0.7999999999999999,
+ "output_per_million": 4.0,
+ "cached_input_per_million": 0.08
}
}
},
"metadata": {
- "description": "Aion-1.0-Mini 32B parameter model is a distilled version of the DeepSeek-R1 model, designed for strong performance in reasoning domains such as mathematics, coding, and logic. It is a modified variant of a FuseAI model that outperforms R1-Distill-Qwen-32B and R1-Distill-Llama-70B, with benchmark results available on its [Hugging Face page](https://huggingface.co/FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview), independently replicated for verification.",
+ "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Claude",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 32768,
- "is_moderated": false
+ "context_length": 200000,
+ "max_completion_tokens": 8192,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
"max_tokens",
- "reasoning",
+ "stop",
"temperature",
+ "tool_choice",
+ "tools",
+ "top_k",
"top_p"
]
}
},
{
- "id": "aion-labs/aion-rp-llama-3.1-8b",
- "name": "AionLabs: Aion-RP 1.0 (8B)",
+ "id": "anthropic/claude-3.5-haiku-20241022",
+ "name": "Anthropic: Claude 3.5 Haiku (2024-10-22)",
"provider": "openrouter",
- "family": "aion-labs",
- "created_at": "2025-02-04 20:18:38 +0100",
- "context_window": 32768,
- "max_output_tokens": 32768,
+ "family": "anthropic",
+ "created_at": "2024-11-04 11:00:00 +1100",
+ "context_window": 200000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.19999999999999998
+ "input_per_million": 0.7999999999999999,
+ "output_per_million": 4.0,
+ "cached_input_per_million": 0.08
}
}
},
"metadata": {
- "description": "Aion-RP-Llama-3.1-8B ranks the highest in the character evaluation portion of the RPBench-Auto benchmark, a roleplaying-specific variant of Arena-Hard-Auto, where LLMs evaluate each other’s responses. It is a fine-tuned base model rather than an instruct model, designed to produce more natural and varied writing.",
+ "description": "Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.\n\nIt does not support image inputs.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Claude",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 32768,
+ "context_length": 200000,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"max_tokens",
+ "stop",
"temperature",
+ "tool_choice",
+ "tools",
+ "top_k",
"top_p"
]
}
},
{
- "id": "alfredpros/codellama-7b-instruct-solidity",
- "name": "AlfredPros: CodeLLaMa 7B Instruct Solidity",
+ "id": "anthropic/claude-3.5-sonnet",
+ "name": "Anthropic: Claude 3.5 Sonnet",
"provider": "openrouter",
- "family": "alfredpros",
- "created_at": "2025-04-14 16:44:34 +0200",
- "context_window": 8192,
+ "family": "anthropic",
+ "created_at": "2024-10-22 11:00:00 +1100",
+ "context_window": 200000,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7,
- "output_per_million": 1.1
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.3
}
}
},
"metadata": {
- "description": "A finetuned 7 billion parameters Code LLaMA - Instruct model to generate Solidity smart contract using 4-bit QLoRA finetuning provided by PEFT library.",
+ "description": "New Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "alpaca"
+ "tokenizer": "Claude",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 8192,
+ "context_length": 200000,
"max_completion_tokens": 8192,
- "is_moderated": false
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "alpindale/goliath-120b",
- "name": "Goliath 120B",
+ "id": "anthropic/claude-3.5-sonnet-20240620",
+ "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20)",
"provider": "openrouter",
- "family": "alpindale",
- "created_at": "2023-11-10 01:00:00 +0100",
- "context_window": 6144,
- "max_output_tokens": 512,
+ "family": "anthropic",
+ "created_at": "2024-06-20 10:00:00 +1000",
+ "context_window": 200000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
@@ -10622,67 +10933,63 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 4.0,
- "output_per_million": 5.5
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.3
}
}
},
"metadata": {
- "description": "A large LLM created by combining two fine-tuned Llama 70B models into one 120B model. Combines Xwin and Euryale.\n\nCredits to\n- [@chargoddard](https://huggingface.co/chargoddard) for developing the framework used to merge the model - [mergekit](https://github.com/cg123/mergekit).\n- [@Undi95](https://huggingface.co/Undi95) for helping with the merge ratios.\n\n#merge",
+ "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama2",
- "instruct_type": "airoboros"
+ "tokenizer": "Claude",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 6144,
- "max_completion_tokens": 512,
- "is_moderated": false
+ "context_length": 200000,
+ "max_completion_tokens": 8192,
+ "is_moderated": true
},
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "response_format",
- "seed",
+ "per_request_limits": null,
+ "supported_parameters": [
+ "max_tokens",
"stop",
- "structured_outputs",
"temperature",
- "top_a",
+ "tool_choice",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "amazon/nova-lite-v1",
- "name": "Amazon: Nova Lite 1.0",
+ "id": "anthropic/claude-3.7-sonnet",
+ "name": "Anthropic: Claude 3.7 Sonnet",
"provider": "openrouter",
- "family": "amazon",
- "created_at": "2024-12-05 23:22:43 +0100",
- "context_window": 300000,
- "max_output_tokens": 5120,
+ "family": "anthropic",
+ "created_at": "2025-02-25 05:35:10 +1100",
+ "context_window": 200000,
+ "max_output_tokens": 64000,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text",
- "image"
+ "image",
+ "file"
],
"output": [
"text"
@@ -10695,35 +11002,40 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.06,
- "output_per_million": 0.24
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.3
}
}
},
"metadata": {
- "description": "Amazon Nova Lite 1.0 is a very low-cost multimodal model from Amazon that focused on fast processing of image, video, and text inputs to generate text output. Amazon Nova Lite can handle real-time customer interactions, document analysis, and visual question-answering tasks with high accuracy.\n\nWith an input context of 300K tokens, it can analyze multiple images or up to 30 minutes of video in a single input.",
+ "description": "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
"text",
- "image"
+ "image",
+ "file"
],
"output_modalities": [
"text"
],
- "tokenizer": "Nova",
+ "tokenizer": "Claude",
"instruct_type": null
},
"top_provider": {
- "context_length": 300000,
- "max_completion_tokens": 5120,
- "is_moderated": true
+ "context_length": 200000,
+ "max_completion_tokens": 64000,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "include_reasoning",
"max_tokens",
+ "reasoning",
"stop",
"temperature",
+ "tool_choice",
"tools",
"top_k",
"top_p"
@@ -10731,17 +11043,19 @@
}
},
{
- "id": "amazon/nova-micro-v1",
- "name": "Amazon: Nova Micro 1.0",
+ "id": "anthropic/claude-3.7-sonnet:thinking",
+ "name": "Anthropic: Claude 3.7 Sonnet (thinking)",
"provider": "openrouter",
- "family": "amazon",
- "created_at": "2024-12-05 23:20:37 +0100",
- "context_window": 128000,
- "max_output_tokens": 5120,
+ "family": "anthropic",
+ "created_at": "2025-02-25 05:35:10 +1100",
+ "context_window": 200000,
+ "max_output_tokens": 64000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
@@ -10754,53 +11068,58 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.035,
- "output_per_million": 0.14
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.3
}
}
},
"metadata": {
- "description": "Amazon Nova Micro 1.0 is a text-only model that delivers the lowest latency responses in the Amazon Nova family of models at a very low cost. With a context length of 128K tokens and optimized for speed and cost, Amazon Nova Micro excels at tasks such as text summarization, translation, content classification, interactive chat, and brainstorming. It has simple mathematical reasoning and coding abilities.",
+ "description": "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
],
- "tokenizer": "Nova",
+ "tokenizer": "Claude",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 5120,
- "is_moderated": true
+ "context_length": 200000,
+ "max_completion_tokens": 64000,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "include_reasoning",
"max_tokens",
+ "reasoning",
"stop",
"temperature",
- "tools",
- "top_k",
- "top_p"
+ "tool_choice",
+ "tools"
]
}
},
{
- "id": "amazon/nova-pro-v1",
- "name": "Amazon: Nova Pro 1.0",
+ "id": "anthropic/claude-opus-4",
+ "name": "Anthropic: Claude Opus 4",
"provider": "openrouter",
- "family": "amazon",
- "created_at": "2024-12-05 23:05:03 +0100",
- "context_window": 300000,
- "max_output_tokens": 5120,
+ "family": "anthropic",
+ "created_at": "2025-05-23 02:27:25 +1000",
+ "context_window": 200000,
+ "max_output_tokens": 32000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text",
- "image"
+ "file"
],
"output": [
"text"
@@ -10813,35 +11132,40 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7999999999999999,
- "output_per_million": 3.1999999999999997
+ "input_per_million": 15.0,
+ "output_per_million": 75.0,
+ "cached_input_per_million": 1.5
}
}
},
"metadata": {
- "description": "Amazon Nova Pro 1.0 is a capable multimodal model from Amazon focused on providing a combination of accuracy, speed, and cost for a wide range of tasks. As of December 2024, it achieves state-of-the-art performance on key benchmarks including visual question answering (TextVQA) and video understanding (VATEX).\n\nAmazon Nova Pro demonstrates strong capabilities in processing both visual and textual information and at analyzing financial documents.\n\n**NOTE**: Video input is not supported at this time.",
+ "description": "Claude Opus 4 is benchmarked as the world’s best coding model, at time of release, bringing sustained performance on complex, long-running tasks and agent workflows. It sets new benchmarks in software engineering, achieving leading results on SWE-bench (72.5%) and Terminal-bench (43.2%). Opus 4 supports extended, agentic workflows, handling thousands of task steps continuously for hours without degradation. \n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-4)",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
+ "image",
"text",
- "image"
+ "file"
],
"output_modalities": [
"text"
],
- "tokenizer": "Nova",
+ "tokenizer": "Claude",
"instruct_type": null
},
"top_provider": {
- "context_length": 300000,
- "max_completion_tokens": 5120,
+ "context_length": 200000,
+ "max_completion_tokens": 32000,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
+ "include_reasoning",
"max_tokens",
+ "reasoning",
"stop",
"temperature",
+ "tool_choice",
"tools",
"top_k",
"top_p"
@@ -10849,17 +11173,19 @@
}
},
{
- "id": "anthracite-org/magnum-v2-72b",
- "name": "Magnum v2 72B",
+ "id": "anthropic/claude-opus-4.1",
+ "name": "Anthropic: Claude Opus 4.1",
"provider": "openrouter",
- "family": "anthracite-org",
- "created_at": "2024-09-30 02:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "anthropic",
+ "created_at": "2025-08-06 02:33:11 +1000",
+ "context_window": 200000,
+ "max_output_tokens": 32000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output": [
"text"
@@ -10867,62 +11193,63 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 3.0
+ "input_per_million": 15.0,
+ "output_per_million": 75.0,
+ "cached_input_per_million": 1.5
}
}
},
"metadata": {
- "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the seventh in a family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.",
+ "description": "Claude Opus 4.1 is an updated version of Anthropic’s flagship model, offering improved performance in coding, reasoning, and agentic tasks. It achieves 74.5% on SWE-bench Verified and shows notable gains in multi-file code refactoring, debugging precision, and detail-oriented reasoning. The model supports extended thinking up to 64K tokens and is optimized for tasks involving research, data analysis, and tool-assisted reasoning.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "chatml"
+ "tokenizer": "Claude",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
- "is_moderated": false
+ "context_length": 200000,
+ "max_completion_tokens": 32000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
+ "include_reasoning",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "seed",
+ "reasoning",
"stop",
"temperature",
- "top_k",
- "top_p"
+ "tool_choice",
+ "tools"
]
}
},
{
- "id": "anthracite-org/magnum-v4-72b",
- "name": "Magnum v4 72B",
+ "id": "anthropic/claude-sonnet-4",
+ "name": "Anthropic: Claude Sonnet 4",
"provider": "openrouter",
- "family": "anthracite-org",
- "created_at": "2024-10-22 02:00:00 +0200",
- "context_window": 16384,
- "max_output_tokens": 1024,
+ "family": "anthropic",
+ "created_at": "2025-05-23 02:12:51 +1000",
+ "context_window": 200000,
+ "max_output_tokens": 64000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output": [
"text"
@@ -10930,64 +11257,63 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 5.0
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.3
}
}
},
"metadata": {
- "description": "This is a series of models designed to replicate the prose quality of the Claude 3 models, specifically Sonnet(https://openrouter.ai/anthropic/claude-3.5-sonnet) and Opus(https://openrouter.ai/anthropic/claude-3-opus).\n\nThe model is fine-tuned on top of [Qwen2.5 72B](https://openrouter.ai/qwen/qwen-2.5-72b-instruct).",
+ "description": "Claude Sonnet 4 significantly enhances the capabilities of its predecessor, Sonnet 3.7, excelling in both coding and reasoning tasks with improved precision and controllability. Achieving state-of-the-art performance on SWE-bench (72.7%), Sonnet 4 balances capability and computational efficiency, making it suitable for a broad range of applications from routine coding tasks to complex software development projects. Key enhancements include improved autonomous codebase navigation, reduced error rates in agent-driven workflows, and increased reliability in following intricate instructions. Sonnet 4 is optimized for practical everyday use, providing advanced reasoning capabilities while maintaining efficiency and responsiveness in diverse internal and external scenarios.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-4)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "chatml"
+ "tokenizer": "Claude",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 16384,
- "max_completion_tokens": 1024,
+ "context_length": 200000,
+ "max_completion_tokens": 64000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
+ "include_reasoning",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "seed",
+ "reasoning",
"stop",
"temperature",
- "top_a",
+ "tool_choice",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "anthropic/claude-3-haiku",
- "name": "Anthropic: Claude 3 Haiku",
+ "id": "arcee-ai/coder-large",
+ "name": "Arcee AI: Coder Large",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2024-03-13 01:00:00 +0100",
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "family": "arcee-ai",
+ "created_at": "2025-05-06 06:57:43 +1000",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -10995,61 +11321,61 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "output_per_million": 1.25,
- "cached_input_per_million": 0.03
+ "input_per_million": 0.5,
+ "output_per_million": 0.7999999999999999
}
}
},
"metadata": {
- "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal",
+ "description": "Coder‑Large is a 32 B‑parameter offspring of Qwen 2.5‑Instruct that has been further trained on permissively‑licensed GitHub, CodeSearchNet and synthetic bug‑fix corpora. It supports a 32k context window, enabling multi‑file refactoring or long diff review in a single call, and understands 30‑plus programming languages with special attention to TypeScript, Go and Terraform. Internal benchmarks show 5–8 pt gains over CodeLlama‑34 B‑Python on HumanEval and competitive BugFix scores thanks to a reinforcement pass that rewards compilable output. The model emits structured explanations alongside code blocks by default, making it suitable for educational tooling as well as production copilot scenarios. Cost‑wise, Together AI prices it well below proprietary incumbents, so teams can scale interactive coding without runaway spend. ",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 4096,
- "is_moderated": true
+ "context_length": 32768,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "anthropic/claude-3-opus",
- "name": "Anthropic: Claude 3 Opus",
+ "id": "arcee-ai/maestro-reasoning",
+ "name": "Arcee AI: Maestro Reasoning",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2024-03-05 01:00:00 +0100",
- "context_window": 200000,
- "max_output_tokens": 4096,
+ "family": "arcee-ai",
+ "created_at": "2025-05-06 07:41:09 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 32000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -11057,61 +11383,62 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 75.0,
- "cached_input_per_million": 1.5
+ "input_per_million": 0.8999999999999999,
+ "output_per_million": 3.3000000000000003
}
}
},
"metadata": {
- "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal",
+ "description": "Maestro Reasoning is Arcee's flagship analysis model: a 32 B‑parameter derivative of Qwen 2.5‑32 B tuned with DPO and chain‑of‑thought RL for step‑by‑step logic. Compared to the earlier 7 B preview, the production 32 B release widens the context window to 128 k tokens and doubles pass‑rate on MATH and GSM‑8K, while also lifting code completion accuracy. Its instruction style encourages structured \"thought → answer\" traces that can be parsed or hidden according to user preference. That transparency pairs well with audit‑focused industries like finance or healthcare where seeing the reasoning path matters. In Arcee Conductor, Maestro is automatically selected for complex, multi‑constraint queries that smaller SLMs bounce. ",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 4096,
- "is_moderated": true
+ "context_length": 131072,
+ "max_completion_tokens": 32000,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "anthropic/claude-3.5-haiku",
- "name": "Anthropic: Claude 3.5 Haiku",
+ "id": "arcee-ai/spotlight",
+ "name": "Arcee AI: Spotlight",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2024-11-04 01:00:00 +0100",
- "context_window": 200000,
- "max_output_tokens": 8192,
+ "family": "arcee-ai",
+ "created_at": "2025-05-06 07:45:52 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 65537,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "image",
+ "text"
],
"output": [
"text"
@@ -11119,62 +11446,62 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7999999999999999,
- "output_per_million": 4.0,
- "cached_input_per_million": 0.08
+ "input_per_million": 0.18,
+ "output_per_million": 0.18
}
}
},
"metadata": {
- "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).",
+ "description": "Spotlight is a 7‑billion‑parameter vision‑language model derived from Qwen 2.5‑VL and fine‑tuned by Arcee AI for tight image‑text grounding tasks. It offers a 32 k‑token context window, enabling rich multimodal conversations that combine lengthy documents with one or more images. Training emphasized fast inference on consumer GPUs while retaining strong captioning, visual‐question‑answering, and diagram‑analysis accuracy. As a result, Spotlight slots neatly into agent workflows where screenshots, charts or UI mock‑ups need to be interpreted on the fly. Early benchmarks show it matching or out‑scoring larger VLMs such as LLaVA‑1.6 13 B on popular VQA and POPE alignment tests. ",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "text",
- "image"
+ "image",
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 8192,
- "is_moderated": true
+ "context_length": 131072,
+ "max_completion_tokens": 65537,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "anthropic/claude-3.5-haiku-20241022",
- "name": "Anthropic: Claude 3.5 Haiku (2024-10-22)",
+ "id": "arcee-ai/virtuoso-large",
+ "name": "Arcee AI: Virtuoso Large",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2024-11-04 01:00:00 +0100",
- "context_window": 200000,
- "max_output_tokens": 8192,
+ "family": "arcee-ai",
+ "created_at": "2025-05-06 07:01:25 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 64000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
@@ -11182,40 +11509,43 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7999999999999999,
- "output_per_million": 4.0,
- "cached_input_per_million": 0.08
+ "input_per_million": 0.75,
+ "output_per_million": 1.2
}
}
},
"metadata": {
- "description": "Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.\n\nIt does not support image inputs.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)",
+ "description": "Virtuoso‑Large is Arcee's top‑tier general‑purpose LLM at 72 B parameters, tuned to tackle cross‑domain reasoning, creative writing and enterprise QA. Unlike many 70 B peers, it retains the 128 k context inherited from Qwen 2.5, letting it ingest books, codebases or financial filings wholesale. Training blended DeepSeek R1 distillation, multi‑epoch supervised fine‑tuning and a final DPO/RLHF alignment stage, yielding strong performance on BIG‑Bench‑Hard, GSM‑8K and long‑context Needle‑In‑Haystack tests. Enterprises use Virtuoso‑Large as the \"fallback\" brain in Conductor pipelines when other SLMs flag low confidence. Despite its size, aggressive KV‑cache optimizations keep first‑token latency in the low‑second range on 8× H100 nodes, making it a practical production‑grade powerhouse.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 8192,
+ "context_length": 131072,
+ "max_completion_tokens": 64000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"stop",
"temperature",
"tool_choice",
@@ -11226,19 +11556,17 @@
}
},
{
- "id": "anthropic/claude-3.5-sonnet",
- "name": "Anthropic: Claude 3.5 Sonnet",
+ "id": "arliai/qwq-32b-arliai-rpr-v1",
+ "name": "ArliAI: QwQ 32B RpR v1",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2024-10-22 02:00:00 +0200",
- "context_window": 200000,
- "max_output_tokens": 8192,
+ "family": "arliai",
+ "created_at": "2025-04-14 00:53:02 +1000",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
@@ -11246,63 +11574,66 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0,
- "cached_input_per_million": 0.3
+ "input_per_million": 0.01,
+ "output_per_million": 0.0400032
}
}
},
"metadata": {
- "description": "New Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal",
+ "description": "QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself.\n\nThe model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
- "instruct_type": null
+ "tokenizer": "Other",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 8192,
- "is_moderated": true
+ "context_length": 32768,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "reasoning",
+ "repetition_penalty",
+ "seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "anthropic/claude-3.5-sonnet-20240620",
- "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20)",
+ "id": "arliai/qwq-32b-arliai-rpr-v1:free",
+ "name": "ArliAI: QwQ 32B RpR v1 (free)",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2024-06-20 02:00:00 +0200",
- "context_window": 200000,
- "max_output_tokens": 8192,
+ "family": "arliai",
+ "created_at": "2025-04-14 00:53:02 +1000",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
@@ -11310,63 +11641,59 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0,
- "cached_input_per_million": 0.3
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal",
+ "description": "QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself.\n\nThe model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
- "instruct_type": null
+ "tokenizer": "Other",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 8192,
- "is_moderated": true
+ "context_length": 32768,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "reasoning",
+ "repetition_penalty",
+ "seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "anthropic/claude-3.7-sonnet",
- "name": "Anthropic: Claude 3.7 Sonnet",
+ "id": "baidu/ernie-4.5-21b-a3b",
+ "name": "Baidu: ERNIE 4.5 21B A3B",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2025-02-24 19:35:10 +0100",
- "context_window": 200000,
- "max_output_tokens": 64000,
+ "family": "baidu",
+ "created_at": "2025-08-13 07:29:27 +1000",
+ "context_window": 120000,
+ "max_output_tokens": 8000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
@@ -11374,65 +11701,62 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0,
- "cached_input_per_million": 0.3
+ "input_per_million": 0.07,
+ "output_per_million": 0.28
}
}
},
"metadata": {
- "description": "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)",
+ "description": "A sophisticated text-based Mixture-of-Experts (MoE) model featuring 21B total parameters with 3B activated per token, delivering exceptional multimodal understanding and generation through heterogeneous MoE structures and modality-isolated routing. Supporting an extensive 131K token context length, the model achieves efficient inference via multi-expert parallel collaboration and quantization, while advanced post-training techniques including SFT, DPO, and UPO ensure optimized performance across diverse applications with specialized routing and balancing losses for superior task handling.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 64000,
+ "context_length": 120000,
+ "max_completion_tokens": 8000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
- "reasoning",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "anthropic/claude-3.7-sonnet:thinking",
- "name": "Anthropic: Claude 3.7 Sonnet (thinking)",
+ "id": "baidu/ernie-4.5-300b-a47b",
+ "name": "Baidu: ERNIE 4.5 300B A47B ",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2025-02-24 19:35:10 +0100",
- "context_window": 200000,
- "max_output_tokens": 64000,
+ "family": "baidu",
+ "created_at": "2025-07-01 02:15:39 +1000",
+ "context_window": 123000,
+ "max_output_tokens": 12000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
@@ -11440,63 +11764,63 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0,
- "cached_input_per_million": 0.3
+ "input_per_million": 0.28,
+ "output_per_million": 1.1
}
}
},
"metadata": {
- "description": "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)",
+ "description": "ERNIE-4.5-300B-A47B is a 300B parameter Mixture-of-Experts (MoE) language model developed by Baidu as part of the ERNIE 4.5 series. It activates 47B parameters per token and supports text generation in both English and Chinese. Optimized for high-throughput inference and efficient scaling, it uses a heterogeneous MoE structure with advanced routing and quantization strategies, including FP8 and 2-bit formats. This version is fine-tuned for language-only tasks and supports reasoning, tool parameters, and extended context lengths up to 131k tokens. Suitable for general-purpose LLM applications with high reasoning and throughput demands.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 64000,
+ "context_length": 123000,
+ "max_completion_tokens": 12000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
- "reasoning",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "seed",
"stop",
"temperature",
- "tool_choice",
- "tools"
+ "top_k",
+ "top_p"
]
}
},
{
- "id": "anthropic/claude-opus-4",
- "name": "Anthropic: Claude Opus 4",
+ "id": "baidu/ernie-4.5-vl-28b-a3b",
+ "name": "Baidu: ERNIE 4.5 VL 28B A3B",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2025-05-22 18:27:25 +0200",
- "context_window": 200000,
- "max_output_tokens": 32000,
+ "family": "baidu",
+ "created_at": "2025-08-13 07:07:16 +1000",
+ "context_window": 30000,
+ "max_output_tokens": 8000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text",
- "file"
+ "image"
],
"output": [
"text"
@@ -11504,65 +11828,66 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 75.0,
- "cached_input_per_million": 1.5
+ "input_per_million": 0.14,
+ "output_per_million": 0.56
}
}
},
"metadata": {
- "description": "Claude Opus 4 is benchmarked as the world’s best coding model, at time of release, bringing sustained performance on complex, long-running tasks and agent workflows. It sets new benchmarks in software engineering, achieving leading results on SWE-bench (72.5%) and Terminal-bench (43.2%). Opus 4 supports extended, agentic workflows, handling thousands of task steps continuously for hours without degradation. \n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-4)",
+ "description": "A powerful multimodal Mixture-of-Experts chat model featuring 28B total parameters with 3B activated per token, delivering exceptional text and vision understanding through its innovative heterogeneous MoE structure with modality-isolated routing. Built with scaling-efficient infrastructure for high-throughput training and inference, the model leverages advanced post-training techniques including SFT, DPO, and UPO for optimized performance, while supporting an impressive 131K context length and RLVR alignment for superior cross-modal reasoning and generation capabilities.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "image",
"text",
- "file"
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 32000,
- "is_moderated": true
+ "context_length": 30000,
+ "max_completion_tokens": 8000,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"include_reasoning",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
"reasoning",
+ "repetition_penalty",
+ "seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "anthropic/claude-opus-4.1",
- "name": "Anthropic: Claude Opus 4.1",
+ "id": "baidu/ernie-4.5-vl-424b-a47b",
+ "name": "Baidu: ERNIE 4.5 VL 424B A47B ",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2025-08-05 18:33:11 +0200",
- "context_window": 200000,
- "max_output_tokens": 32000,
+ "family": "baidu",
+ "created_at": "2025-07-01 02:28:23 +1000",
+ "context_window": 123000,
+ "max_output_tokens": 16000,
"knowledge_cutoff": null,
"modalities": {
"input": [
"image",
- "text",
- "file"
+ "text"
],
"output": [
"text"
@@ -11570,121 +11895,122 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 75.0,
- "cached_input_per_million": 1.5
+ "input_per_million": 0.42,
+ "output_per_million": 1.25
}
}
},
"metadata": {
- "description": "Claude Opus 4.1 is an updated version of Anthropic’s flagship model, offering improved performance in coding, reasoning, and agentic tasks. It achieves 74.5% on SWE-bench Verified and shows notable gains in multi-file code refactoring, debugging precision, and detail-oriented reasoning. The model supports extended thinking up to 64K tokens and is optimized for tasks involving research, data analysis, and tool-assisted reasoning.",
+ "description": "ERNIE-4.5-VL-424B-A47B is a multimodal Mixture-of-Experts (MoE) model from Baidu’s ERNIE 4.5 series, featuring 424B total parameters with 47B active per token. It is trained jointly on text and image data using a heterogeneous MoE architecture and modality-isolated routing to enable high-fidelity cross-modal reasoning, image understanding, and long-context generation (up to 131k tokens). Fine-tuned with techniques like SFT, DPO, UPO, and RLVR, this model supports both “thinking” and non-thinking inference modes. Designed for vision-language tasks in English and Chinese, it is optimized for efficient scaling and can operate under 4-bit/8-bit quantization.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
"image",
- "text",
- "file"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 32000,
- "is_moderated": true
+ "context_length": 123000,
+ "max_completion_tokens": 16000,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"include_reasoning",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
"reasoning",
+ "repetition_penalty",
+ "seed",
"stop",
"temperature",
- "tool_choice",
- "tools"
+ "top_k",
+ "top_p"
]
}
},
{
- "id": "anthropic/claude-sonnet-4",
- "name": "Anthropic: Claude Sonnet 4",
+ "id": "bytedance/ui-tars-1.5-7b",
+ "name": "Bytedance: UI-TARS 7B ",
"provider": "openrouter",
- "family": "anthropic",
- "created_at": "2025-05-22 18:12:51 +0200",
- "context_window": 200000,
- "max_output_tokens": 64000,
+ "family": "bytedance",
+ "created_at": "2025-07-23 03:24:16 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
"image",
- "text",
- "file"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0,
- "cached_input_per_million": 0.3
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.19999999999999998
}
}
},
"metadata": {
- "description": "Claude Sonnet 4 significantly enhances the capabilities of its predecessor, Sonnet 3.7, excelling in both coding and reasoning tasks with improved precision and controllability. Achieving state-of-the-art performance on SWE-bench (72.7%), Sonnet 4 balances capability and computational efficiency, making it suitable for a broad range of applications from routine coding tasks to complex software development projects. Key enhancements include improved autonomous codebase navigation, reduced error rates in agent-driven workflows, and increased reliability in following intricate instructions. Sonnet 4 is optimized for practical everyday use, providing advanced reasoning capabilities while maintaining efficiency and responsiveness in diverse internal and external scenarios.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-4)",
+ "description": "UI-TARS-1.5 is a multimodal vision-language agent optimized for GUI-based environments, including desktop interfaces, web browsers, mobile systems, and games. Built by ByteDance, it builds upon the UI-TARS framework with reinforcement learning-based reasoning, enabling robust action planning and execution across virtual interfaces.\n\nThis model achieves state-of-the-art results on a range of interactive and grounding benchmarks, including OSworld, WebVoyager, AndroidWorld, and ScreenSpot. It also demonstrates perfect task completion across diverse Poki games and outperforms prior models in Minecraft agent tasks. UI-TARS-1.5 supports thought decomposition during inference and shows strong scaling across variants, with the 1.5 version notably exceeding the performance of earlier 72B and 7B checkpoints.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
"image",
- "text",
- "file"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Claude",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 64000,
+ "context_length": 128000,
+ "max_completion_tokens": 2048,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
"max_tokens",
- "reasoning",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "arcee-ai/coder-large",
- "name": "Arcee AI: Coder Large",
+ "id": "cognitivecomputations/dolphin-mistral-24b-venice-edition:free",
+ "name": "Venice: Uncensored (free)",
"provider": "openrouter",
- "family": "arcee-ai",
- "created_at": "2025-05-05 22:57:43 +0200",
+ "family": "cognitivecomputations",
+ "created_at": "2025-07-10 07:02:46 +1000",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -11698,18 +12024,11 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.5,
- "output_per_million": 0.7999999999999999
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Coder‑Large is a 32 B‑parameter offspring of Qwen 2.5‑Instruct that has been further trained on permissively‑licensed GitHub, CodeSearchNet and synthetic bug‑fix corpora. It supports a 32k context window, enabling multi‑file refactoring or long diff review in a single call, and understands 30‑plus programming languages with special attention to TypeScript, Go and Terraform. Internal benchmarks show 5–8 pt gains over CodeLlama‑34 B‑Python on HumanEval and competitive BugFix scores thanks to a reinforcement pass that rewards compilable output. The model emits structured explanations alongside code blocks by default, making it suitable for educational tooling as well as production copilot scenarios. Cost‑wise, Together AI prices it well below proprietary incumbents, so teams can scale interactive coding without runaway spend. ",
+ "description": "Venice Uncensored Dolphin Mistral 24B Venice Edition is a fine-tuned variant of Mistral-Small-24B-Instruct-2501, developed by dphn.ai in collaboration with Venice.ai. This model is designed as an “uncensored” instruct-tuned LLM, preserving user control over alignment, system prompts, and behavior. Intended for advanced and unrestricted use cases, Venice Uncensored emphasizes steerability and transparent behavior, removing default safety and alignment layers typically found in mainstream assistant models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -11729,12 +12048,11 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"stop",
+ "structured_outputs",
"temperature",
"top_k",
"top_p"
@@ -11742,13 +12060,13 @@
}
},
{
- "id": "arcee-ai/maestro-reasoning",
- "name": "Arcee AI: Maestro Reasoning",
+ "id": "cognitivecomputations/dolphin-mixtral-8x22b",
+ "name": "Dolphin 2.9.2 Mixtral 8x22B 🐬",
"provider": "openrouter",
- "family": "arcee-ai",
- "created_at": "2025-05-05 23:41:09 +0200",
- "context_window": 131072,
- "max_output_tokens": 32000,
+ "family": "cognitivecomputations",
+ "created_at": "2024-06-08 10:00:00 +1000",
+ "context_window": 16000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -11766,12 +12084,12 @@
"text_tokens": {
"standard": {
"input_per_million": 0.8999999999999999,
- "output_per_million": 3.3000000000000003
+ "output_per_million": 0.8999999999999999
}
}
},
"metadata": {
- "description": "Maestro Reasoning is Arcee's flagship analysis model: a 32 B‑parameter derivative of Qwen 2.5‑32 B tuned with DPO and chain‑of‑thought RL for step‑by‑step logic. Compared to the earlier 7 B preview, the production 32 B release widens the context window to 128 k tokens and doubles pass‑rate on MATH and GSM‑8K, while also lifting code completion accuracy. Its instruction style encourages structured \"thought → answer\" traces that can be parsed or hidden according to user preference. That transparency pairs well with audit‑focused industries like finance or healthcare where seeing the reasoning path matters. In Arcee Conductor, Maestro is automatically selected for complex, multi‑constraint queries that smaller SLMs bounce. ",
+ "description": "Dolphin 2.9 is designed for instruction following, conversational, and coding. This model is a finetune of [Mixtral 8x22B Instruct](/models/mistralai/mixtral-8x22b-instruct). It features a 64k context length and was fine-tuned with a 16k sequence length using ChatML templates.\n\nThis model is a successor to [Dolphin Mixtral 8x7B](/models/cognitivecomputations/dolphin-mixtral-8x7b).\n\nThe model is uncensored and is stripped of alignment and bias. It requires an external alignment layer for ethical use. Users are cautioned to use this highly compliant model responsibly, as detailed in a blog post about uncensored models at [erichartford.com/uncensored-models](https://erichartford.com/uncensored-models).\n\n#moe #uncensored",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -11780,12 +12098,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Mistral",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 32000,
+ "context_length": 16000,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
@@ -11796,6 +12114,7 @@
"min_p",
"presence_penalty",
"repetition_penalty",
+ "seed",
"stop",
"temperature",
"top_k",
@@ -11804,17 +12123,16 @@
}
},
{
- "id": "arcee-ai/spotlight",
- "name": "Arcee AI: Spotlight",
+ "id": "cognitivecomputations/dolphin3.0-mistral-24b",
+ "name": "Dolphin3.0 Mistral 24B",
"provider": "openrouter",
- "family": "arcee-ai",
- "created_at": "2025-05-05 23:45:52 +0200",
- "context_window": 131072,
- "max_output_tokens": 65537,
+ "family": "cognitivecomputations",
+ "created_at": "2025-02-14 02:53:39 +1100",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -11828,17 +12146,16 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.18,
- "output_per_million": 0.18
+ "input_per_million": 0.037022,
+ "output_per_million": 0.14816
}
}
},
"metadata": {
- "description": "Spotlight is a 7‑billion‑parameter vision‑language model derived from Qwen 2.5‑VL and fine‑tuned by Arcee AI for tight image‑text grounding tasks. It offers a 32 k‑token context window, enabling rich multimodal conversations that combine lengthy documents with one or more images. Training emphasized fast inference on consumer GPUs while retaining strong captioning, visual‐question‑answering, and diagram‑analysis accuracy. As a result, Spotlight slots neatly into agent workflows where screenshots, charts or UI mock‑ups need to be interpreted on the fly. Early benchmarks show it matching or out‑scoring larger VLMs such as LLaVA‑1.6 13 B on popular VQA and POPE alignment tests. ",
+ "description": "Dolphin 3.0 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nDolphin aims to be a general purpose instruct model, similar to the models behind ChatGPT, Claude, Gemini. \n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
@@ -11848,33 +12165,36 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 65537,
+ "context_length": 32768,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
+ "seed",
"stop",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "arcee-ai/virtuoso-large",
- "name": "Arcee AI: Virtuoso Large",
+ "id": "cognitivecomputations/dolphin3.0-mistral-24b:free",
+ "name": "Dolphin3.0 Mistral 24B (free)",
"provider": "openrouter",
- "family": "arcee-ai",
- "created_at": "2025-05-05 23:01:25 +0200",
- "context_window": 131072,
- "max_output_tokens": 64000,
+ "family": "cognitivecomputations",
+ "created_at": "2025-02-14 02:53:39 +1100",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -11886,19 +12206,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.75,
- "output_per_million": 1.2
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Virtuoso‑Large is Arcee's top‑tier general‑purpose LLM at 72 B parameters, tuned to tackle cross‑domain reasoning, creative writing and enterprise QA. Unlike many 70 B peers, it retains the 128 k context inherited from Qwen 2.5, letting it ingest books, codebases or financial filings wholesale. Training blended DeepSeek R1 distillation, multi‑epoch supervised fine‑tuning and a final DPO/RLHF alignment stage, yielding strong performance on BIG‑Bench‑Hard, GSM‑8K and long‑context Needle‑In‑Haystack tests. Enterprises use Virtuoso‑Large as the \"fallback\" brain in Conductor pipelines when other SLMs flag low confidence. Despite its size, aggressive KV‑cache optimizations keep first‑token latency in the low‑second range on 8× H100 nodes, making it a practical production‑grade powerhouse.",
+ "description": "Dolphin 3.0 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nDolphin aims to be a general purpose instruct model, similar to the models behind ChatGPT, Claude, Gemini. \n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -11911,33 +12223,34 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 64000,
+ "context_length": 32768,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
+ "seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "arliai/qwq-32b-arliai-rpr-v1",
- "name": "ArliAI: QwQ 32B RpR v1",
+ "id": "cognitivecomputations/dolphin3.0-r1-mistral-24b",
+ "name": "Dolphin3.0 R1 Mistral 24B",
"provider": "openrouter",
- "family": "arliai",
- "created_at": "2025-04-13 16:53:02 +0200",
+ "family": "cognitivecomputations",
+ "created_at": "2025-02-14 03:01:38 +1100",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -11957,12 +12270,12 @@
"text_tokens": {
"standard": {
"input_per_million": 0.01,
- "output_per_million": 0.0400032
+ "output_per_million": 0.0340768
}
}
},
"metadata": {
- "description": "QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself.\n\nThe model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.",
+ "description": "Dolphin 3.0 R1 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nThe R1 version has been trained for 3 epochs to reason using 800k reasoning traces from the Dolphin-R1 dataset.\n\nDolphin aims to be a general purpose reasoning instruct model, similar to the models behind ChatGPT, Claude, Gemini.\n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12000,11 +12313,11 @@
}
},
{
- "id": "arliai/qwq-32b-arliai-rpr-v1:free",
- "name": "ArliAI: QwQ 32B RpR v1 (free)",
+ "id": "cognitivecomputations/dolphin3.0-r1-mistral-24b:free",
+ "name": "Dolphin3.0 R1 Mistral 24B (free)",
"provider": "openrouter",
- "family": "arliai",
- "created_at": "2025-04-13 16:53:02 +0200",
+ "family": "cognitivecomputations",
+ "created_at": "2025-02-14 03:01:38 +1100",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -12022,7 +12335,7 @@
],
"pricing": {},
"metadata": {
- "description": "QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself.\n\nThe model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.",
+ "description": "Dolphin 3.0 R1 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nThe R1 version has been trained for 3 epochs to reason using 800k reasoning traces from the Dolphin-R1 dataset.\n\nDolphin aims to be a general purpose reasoning instruct model, similar to the models behind ChatGPT, Claude, Gemini.\n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12060,13 +12373,13 @@
}
},
{
- "id": "baidu/ernie-4.5-21b-a3b",
- "name": "Baidu: ERNIE 4.5 21B A3B",
+ "id": "cohere/command",
+ "name": "Cohere: Command",
"provider": "openrouter",
- "family": "baidu",
- "created_at": "2025-08-12 23:29:27 +0200",
- "context_window": 120000,
- "max_output_tokens": 8000,
+ "family": "cohere",
+ "created_at": "2024-03-14 11:00:00 +1100",
+ "context_window": 4096,
+ "max_output_tokens": 4000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12078,18 +12391,18 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.07,
- "output_per_million": 0.28
+ "input_per_million": 1.0,
+ "output_per_million": 2.0
}
}
},
"metadata": {
- "description": "A sophisticated text-based Mixture-of-Experts (MoE) model featuring 21B total parameters with 3B activated per token, delivering exceptional multimodal understanding and generation through heterogeneous MoE structures and modality-isolated routing. Supporting an extensive 131K token context length, the model achieves efficient inference via multi-expert parallel collaboration and quantization, while advanced post-training techniques including SFT, DPO, and UPO ensure optimized performance across diverse applications with specialized routing and balancing losses for superior task handling.",
+ "description": "Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12098,24 +12411,23 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Cohere",
"instruct_type": null
},
"top_provider": {
- "context_length": 120000,
- "max_completion_tokens": 8000,
- "is_moderated": false
+ "context_length": 4096,
+ "max_completion_tokens": 4000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"top_k",
"top_p"
@@ -12123,13 +12435,13 @@
}
},
{
- "id": "baidu/ernie-4.5-300b-a47b",
- "name": "Baidu: ERNIE 4.5 300B A47B ",
+ "id": "cohere/command-a",
+ "name": "Cohere: Command A",
"provider": "openrouter",
- "family": "baidu",
- "created_at": "2025-06-30 18:15:39 +0200",
- "context_window": 123000,
- "max_output_tokens": 12000,
+ "family": "cohere",
+ "created_at": "2025-03-14 06:32:22 +1100",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12141,18 +12453,18 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.28,
- "output_per_million": 1.1
+ "input_per_million": 2.0,
+ "output_per_million": 8.0
}
}
},
"metadata": {
- "description": "ERNIE-4.5-300B-A47B is a 300B parameter Mixture-of-Experts (MoE) language model developed by Baidu as part of the ERNIE 4.5 series. It activates 47B parameters per token and supports text generation in both English and Chinese. Optimized for high-throughput inference and efficient scaling, it uses a heterogeneous MoE structure with advanced routing and quantization strategies, including FP8 and 2-bit formats. This version is fine-tuned for language-only tasks and supports reasoning, tool parameters, and extended context lengths up to 131k tokens. Suitable for general-purpose LLM applications with high reasoning and throughput demands.",
+ "description": "Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.\nCompared to other leading proprietary and open-weights models Command A delivers maximum performance with minimum hardware costs, excelling on business-critical agentic and multilingual tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12165,20 +12477,19 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 123000,
- "max_completion_tokens": 12000,
+ "context_length": 32768,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"top_k",
"top_p"
@@ -12186,18 +12497,17 @@
}
},
{
- "id": "baidu/ernie-4.5-vl-28b-a3b",
- "name": "Baidu: ERNIE 4.5 VL 28B A3B",
+ "id": "cohere/command-r",
+ "name": "Cohere: Command R",
"provider": "openrouter",
- "family": "baidu",
- "created_at": "2025-08-12 23:07:16 +0200",
- "context_window": 30000,
- "max_output_tokens": 8000,
+ "family": "cohere",
+ "created_at": "2024-03-14 11:00:00 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 4000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -12205,65 +12515,62 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.14,
- "output_per_million": 0.56
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
}
}
},
"metadata": {
- "description": "A powerful multimodal Mixture-of-Experts chat model featuring 28B total parameters with 3B activated per token, delivering exceptional text and vision understanding through its innovative heterogeneous MoE structure with modality-isolated routing. Built with scaling-efficient infrastructure for high-throughput training and inference, the model leverages advanced post-training techniques including SFT, DPO, and UPO for optimized performance, while supporting an impressive 131K context length and RLVR alignment for superior cross-modal reasoning and generation capabilities.",
+ "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Cohere",
"instruct_type": null
},
"top_provider": {
- "context_length": 30000,
- "max_completion_tokens": 8000,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 4000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
- "reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "baidu/ernie-4.5-vl-424b-a47b",
- "name": "Baidu: ERNIE 4.5 VL 424B A47B ",
+ "id": "cohere/command-r-03-2024",
+ "name": "Cohere: Command R (03-2024)",
"provider": "openrouter",
- "family": "baidu",
- "created_at": "2025-06-30 18:28:23 +0200",
- "context_window": 123000,
- "max_output_tokens": 16000,
+ "family": "cohere",
+ "created_at": "2024-03-02 12:00:00 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 4000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -12272,65 +12579,126 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.42,
- "output_per_million": 1.25
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
}
}
},
"metadata": {
- "description": "ERNIE-4.5-VL-424B-A47B is a multimodal Mixture-of-Experts (MoE) model from Baidu’s ERNIE 4.5 series, featuring 424B total parameters with 47B active per token. It is trained jointly on text and image data using a heterogeneous MoE architecture and modality-isolated routing to enable high-fidelity cross-modal reasoning, image understanding, and long-context generation (up to 131k tokens). Fine-tuned with techniques like SFT, DPO, UPO, and RLVR, this model supports both “thinking” and non-thinking inference modes. Designed for vision-language tasks in English and Chinese, it is optimized for efficient scaling and can operate under 4-bit/8-bit quantization.",
+ "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Cohere",
"instruct_type": null
},
"top_provider": {
- "context_length": 123000,
- "max_completion_tokens": 16000,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 4000,
+ "is_moderated": true
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "frequency_penalty",
+ "max_tokens",
+ "presence_penalty",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
+ "temperature",
+ "tools",
+ "top_k",
+ "top_p"
+ ]
+ }
+ },
+ {
+ "id": "cohere/command-r-08-2024",
+ "name": "Cohere: Command R (08-2024)",
+ "provider": "openrouter",
+ "family": "cohere",
+ "created_at": "2024-08-30 10:00:00 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 4000,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
+ }
+ }
+ },
+ "metadata": {
+ "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Cohere",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 128000,
+ "max_completion_tokens": 4000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
- "reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "bytedance/ui-tars-1.5-7b",
- "name": "Bytedance: UI-TARS 7B ",
+ "id": "cohere/command-r-plus",
+ "name": "Cohere: Command R+",
"provider": "openrouter",
- "family": "bytedance",
- "created_at": "2025-07-22 19:24:16 +0200",
+ "family": "cohere",
+ "created_at": "2024-04-04 11:00:00 +1100",
"context_window": 128000,
- "max_output_tokens": 2048,
+ "max_output_tokens": 4000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -12338,58 +12706,60 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.19999999999999998
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
}
}
},
"metadata": {
- "description": "UI-TARS-1.5 is a multimodal vision-language agent optimized for GUI-based environments, including desktop interfaces, web browsers, mobile systems, and games. Built by ByteDance, it builds upon the UI-TARS framework with reinforcement learning-based reasoning, enabling robust action planning and execution across virtual interfaces.\n\nThis model achieves state-of-the-art results on a range of interactive and grounding benchmarks, including OSworld, WebVoyager, AndroidWorld, and ScreenSpot. It also demonstrates perfect task completion across diverse Poki games and outperforms prior models in Minecraft agent tasks. UI-TARS-1.5 supports thought decomposition during inference and shows strong scaling across variants, with the 1.5 version notably exceeding the performance of earlier 72B and 7B checkpoints.",
+ "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Cohere",
"instruct_type": null
},
"top_provider": {
"context_length": 128000,
- "max_completion_tokens": 2048,
- "is_moderated": false
+ "max_completion_tokens": 4000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "cognitivecomputations/dolphin-mistral-24b-venice-edition:free",
- "name": "Venice: Uncensored (free)",
+ "id": "cohere/command-r-plus-04-2024",
+ "name": "Cohere: Command R+ (04-2024)",
"provider": "openrouter",
- "family": "cognitivecomputations",
- "created_at": "2025-07-09 23:02:46 +0200",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "cohere",
+ "created_at": "2024-04-02 11:00:00 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 4000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12401,11 +12771,19 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
+ }
+ }
+ },
"metadata": {
- "description": "Venice Uncensored Dolphin Mistral 24B Venice Edition is a fine-tuned variant of Mistral-Small-24B-Instruct-2501, developed by dphn.ai in collaboration with Venice.ai. This model is designed as an “uncensored” instruct-tuned LLM, preserving user control over alignment, system prompts, and behavior. Intended for advanced and unrestricted use cases, Venice Uncensored emphasizes steerability and transparent behavior, removing default safety and alignment layers typically found in mainstream assistant models.",
+ "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12414,13 +12792,13 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Cohere",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 4000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
@@ -12428,22 +12806,24 @@
"max_tokens",
"presence_penalty",
"response_format",
+ "seed",
"stop",
"structured_outputs",
"temperature",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "cognitivecomputations/dolphin-mixtral-8x22b",
- "name": "Dolphin 2.9.2 Mixtral 8x22B 🐬",
+ "id": "cohere/command-r-plus-08-2024",
+ "name": "Cohere: Command R+ (08-2024)",
"provider": "openrouter",
- "family": "cognitivecomputations",
- "created_at": "2024-06-08 02:00:00 +0200",
- "context_window": 16000,
- "max_output_tokens": 8192,
+ "family": "cohere",
+ "created_at": "2024-08-30 10:00:00 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 4000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12455,18 +12835,19 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.8999999999999999,
- "output_per_million": 0.8999999999999999
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "Dolphin 2.9 is designed for instruction following, conversational, and coding. This model is a finetune of [Mixtral 8x22B Instruct](/models/mistralai/mixtral-8x22b-instruct). It features a 64k context length and was fine-tuned with a 16k sequence length using ChatML templates.\n\nThis model is a successor to [Dolphin Mixtral 8x7B](/models/cognitivecomputations/dolphin-mixtral-8x7b).\n\nThe model is uncensored and is stripped of alignment and bias. It requires an external alignment layer for ethical use. Users are cautioned to use this highly compliant model responsibly, as detailed in a blog post about uncensored models at [erichartford.com/uncensored-models](https://erichartford.com/uncensored-models).\n\n#moe #uncensored",
+ "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12475,38 +12856,38 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
- "instruct_type": "chatml"
+ "tokenizer": "Cohere",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 16000,
- "max_completion_tokens": 8192,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 4000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "cognitivecomputations/dolphin3.0-mistral-24b",
- "name": "Dolphin3.0 Mistral 24B",
+ "id": "cohere/command-r7b-12-2024",
+ "name": "Cohere: Command R7B (12-2024)",
"provider": "openrouter",
- "family": "cognitivecomputations",
- "created_at": "2025-02-13 16:53:39 +0100",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "cohere",
+ "created_at": "2024-12-14 17:35:52 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 4000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12518,18 +12899,18 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.037022,
- "output_per_million": 0.14816
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "description": "Dolphin 3.0 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nDolphin aims to be a general purpose instruct model, similar to the models behind ChatGPT, Claude, Gemini. \n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)",
+ "description": "Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12538,39 +12919,36 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Cohere",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 4000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "cognitivecomputations/dolphin3.0-mistral-24b:free",
- "name": "Dolphin3.0 Mistral 24B (free)",
+ "id": "deepseek/deepseek-chat",
+ "name": "DeepSeek: DeepSeek V3",
"provider": "openrouter",
- "family": "cognitivecomputations",
- "created_at": "2025-02-13 16:53:39 +0100",
- "context_window": 32768,
+ "family": "deepseek",
+ "created_at": "2024-12-27 06:28:40 +1100",
+ "context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -12583,11 +12961,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.1999188,
+ "output_per_million": 0.800064
+ }
+ }
+ },
"metadata": {
- "description": "Dolphin 3.0 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nDolphin aims to be a general purpose instruct model, similar to the models behind ChatGPT, Claude, Gemini. \n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)",
+ "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12596,11 +12983,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "DeepSeek",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 163840,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -12613,9 +13000,13 @@
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -12623,12 +13014,12 @@
}
},
{
- "id": "cognitivecomputations/dolphin3.0-r1-mistral-24b",
- "name": "Dolphin3.0 R1 Mistral 24B",
+ "id": "deepseek/deepseek-chat-v3-0324",
+ "name": "DeepSeek: DeepSeek V3 0324",
"provider": "openrouter",
- "family": "cognitivecomputations",
- "created_at": "2025-02-13 17:01:38 +0100",
- "context_window": 32768,
+ "family": "deepseek",
+ "created_at": "2025-03-25 00:59:15 +1100",
+ "context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -12641,18 +13032,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01,
- "output_per_million": 0.0340768
+ "input_per_million": 0.1999188,
+ "output_per_million": 0.800064
}
}
},
"metadata": {
- "description": "Dolphin 3.0 R1 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nThe R1 version has been trained for 3 epochs to reason using 800k reasoning traces from the Dolphin-R1 dataset.\n\nDolphin aims to be a general purpose reasoning instruct model, similar to the models behind ChatGPT, Claude, Gemini.\n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)",
+ "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12661,28 +13054,30 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "DeepSeek",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 163840,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -12690,12 +13085,12 @@
}
},
{
- "id": "cognitivecomputations/dolphin3.0-r1-mistral-24b:free",
- "name": "Dolphin3.0 R1 Mistral 24B (free)",
+ "id": "deepseek/deepseek-chat-v3-0324:free",
+ "name": "DeepSeek: DeepSeek V3 0324 (free)",
"provider": "openrouter",
- "family": "cognitivecomputations",
- "created_at": "2025-02-13 17:01:38 +0100",
- "context_window": 32768,
+ "family": "deepseek",
+ "created_at": "2025-03-25 00:59:15 +1100",
+ "context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -12708,11 +13103,12 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"predicted_outputs"
],
"pricing": {},
"metadata": {
- "description": "Dolphin 3.0 R1 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nThe R1 version has been trained for 3 epochs to reason using 800k reasoning traces from the Dolphin-R1 dataset.\n\nDolphin aims to be a general purpose reasoning instruct model, similar to the models behind ChatGPT, Claude, Gemini.\n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)",
+ "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12721,28 +13117,28 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "DeepSeek",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 163840,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
"seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -12750,13 +13146,13 @@
}
},
{
- "id": "cohere/command",
- "name": "Cohere: Command",
+ "id": "deepseek/deepseek-chat-v3.1",
+ "name": "DeepSeek: DeepSeek V3.1",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-03-14 01:00:00 +0100",
- "context_window": 4096,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-08-21 22:33:48 +1000",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12768,18 +13164,20 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.0,
- "output_per_million": 2.0
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.7999999999999999
}
}
},
"metadata": {
- "description": "Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)\n\nThe model improves tool use, code generation, and reasoning efficiency, achieving performance comparable to DeepSeek-R1 on difficult benchmarks while responding more quickly. It supports structured tool calling, code agents, and search agents, making it suitable for research, coding, and agentic workflows. \n\nIt succeeds the [DeepSeek V3-0324](/deepseek/deepseek-chat-v3-0324) model and performs well on a variety of tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12788,36 +13186,45 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
- "instruct_type": null
+ "tokenizer": "DeepSeek",
+ "instruct_type": "deepseek-v3.1"
},
"top_provider": {
- "context_length": 4096,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 163840,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "cohere/command-a",
- "name": "Cohere: Command A",
+ "id": "deepseek/deepseek-prover-v2",
+ "name": "DeepSeek: DeepSeek Prover V2",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2025-03-13 20:32:22 +0100",
- "context_window": 32768,
+ "family": "deepseek",
+ "created_at": "2025-04-30 21:38:14 +1000",
+ "context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -12835,13 +13242,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 8.0
+ "input_per_million": 0.5,
+ "output_per_million": 2.1799999999999997
}
}
},
"metadata": {
- "description": "Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.\nCompared to other leading proprietary and open-weights models Command A delivers maximum performance with minimum hardware costs, excelling on business-critical agentic and multilingual tasks.",
+ "description": "DeepSeek Prover V2 is a 671B parameter model, speculated to be geared towards logic and mathematics. Likely an upgrade from [DeepSeek-Prover-V1.5](https://huggingface.co/deepseek-ai/DeepSeek-Prover-V1.5-RL) Not much is known about the model yet, as DeepSeek released it on Hugging Face without an announcement or description.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12850,11 +13257,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "DeepSeek",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 163840,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -12862,11 +13269,12 @@
"supported_parameters": [
"frequency_penalty",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"top_k",
"top_p"
@@ -12874,13 +13282,13 @@
}
},
{
- "id": "cohere/command-r",
- "name": "Cohere: Command R",
+ "id": "deepseek/deepseek-r1",
+ "name": "DeepSeek: R1",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-03-14 01:00:00 +0100",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-01-21 00:51:35 +1100",
+ "context_window": 163840,
+ "max_output_tokens": 163840,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12893,18 +13301,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 2.0
}
}
},
"metadata": {
- "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "DeepSeek R1 is here: Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model & [technical report](https://api-docs.deepseek.com/news/news250120).\n\nMIT licensed: Distill & commercialize freely!",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12913,38 +13322,46 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
- "instruct_type": null
+ "tokenizer": "DeepSeek",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 163840,
+ "max_completion_tokens": 163840,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
"tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "cohere/command-r-03-2024",
- "name": "Cohere: Command R (03-2024)",
+ "id": "deepseek/deepseek-r1-0528",
+ "name": "DeepSeek: R1 0528",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-03-02 02:00:00 +0100",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-05-29 03:59:30 +1000",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12957,18 +13374,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 0.1999188,
+ "output_per_million": 0.800064
}
}
},
"metadata": {
- "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12977,38 +13395,46 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
- "instruct_type": null
+ "tokenizer": "DeepSeek",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 163840,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
"tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "cohere/command-r-08-2024",
- "name": "Cohere: Command R (08-2024)",
+ "id": "deepseek/deepseek-r1-0528-qwen3-8b",
+ "name": "DeepSeek: Deepseek R1 0528 Qwen3 8B",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-08-30 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-05-30 03:09:03 +1000",
+ "context_window": 32000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13020,19 +13446,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6
+ "input_per_million": 0.01,
+ "output_per_million": 0.02
}
}
},
"metadata": {
- "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "DeepSeek-R1-0528 is a lightly upgraded release of DeepSeek R1 that taps more compute and smarter post-training tricks, pushing its reasoning and inference to the brink of flagship models like O3 and Gemini 2.5 Pro.\nIt now tops math, programming, and logic leaderboards, showcasing a step-change in depth-of-thought.\nThe distilled variant, DeepSeek-R1-0528-Qwen3-8B, transfers this chain-of-thought into an 8 B-parameter form, beating standard Qwen3 8B by +10 pp and tying the 235 B “thinking” giant on AIME 2024.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13041,38 +13466,42 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 32000,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "reasoning",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "cohere/command-r-plus",
- "name": "Cohere: Command R+",
+ "id": "deepseek/deepseek-r1-0528-qwen3-8b:free",
+ "name": "DeepSeek: Deepseek R1 0528 Qwen3 8B (free)",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-04-04 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-05-30 03:09:03 +1000",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13084,19 +13513,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "DeepSeek-R1-0528 is a lightly upgraded release of DeepSeek R1 that taps more compute and smarter post-training tricks, pushing its reasoning and inference to the brink of flagship models like O3 and Gemini 2.5 Pro.\nIt now tops math, programming, and logic leaderboards, showcasing a step-change in depth-of-thought.\nThe distilled variant, DeepSeek-R1-0528-Qwen3-8B, transfers this chain-of-thought into an 8 B-parameter form, beating standard Qwen3 8B by +10 pp and tying the 235 B “thinking” giant on AIME 2024.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13105,38 +13526,42 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 131072,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "reasoning",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "cohere/command-r-plus-04-2024",
- "name": "Cohere: Command R+ (04-2024)",
+ "id": "deepseek/deepseek-r1-0528:free",
+ "name": "DeepSeek: R1 0528 (free)",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-04-02 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-05-29 03:59:30 +1000",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13148,19 +13573,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13169,38 +13586,42 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
- "instruct_type": null
+ "tokenizer": "DeepSeek",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 163840,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "reasoning",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "cohere/command-r-plus-08-2024",
- "name": "Cohere: Command R+ (08-2024)",
+ "id": "deepseek/deepseek-r1-distill-llama-70b",
+ "name": "DeepSeek: R1 Distill Llama 70B",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-08-30 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-01-24 07:12:49 +1100",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13213,18 +13634,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 0.025915399999999998,
+ "output_per_million": 0.103712
}
}
},
"metadata": {
- "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13233,38 +13655,45 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 131072,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
+ "tool_choice",
"tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "cohere/command-r7b-12-2024",
- "name": "Cohere: Command R7B (12-2024)",
+ "id": "deepseek/deepseek-r1-distill-llama-70b:free",
+ "name": "DeepSeek: R1 Distill Llama 70B (free)",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-12-14 07:35:52 +0100",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-01-24 07:12:49 +1100",
+ "context_window": 8192,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13276,18 +13705,11 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13296,37 +13718,42 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 8192,
+ "max_completion_tokens": 4096,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "reasoning",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-chat",
- "name": "DeepSeek: DeepSeek V3",
+ "id": "deepseek/deepseek-r1-distill-llama-8b",
+ "name": "DeepSeek: R1 Distill Llama 8B",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2024-12-26 20:28:40 +0100",
- "context_window": 163840,
- "max_output_tokens": null,
+ "created_at": "2025-02-08 01:15:18 +1100",
+ "context_window": 32000,
+ "max_output_tokens": 32000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13338,20 +13765,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1999188,
- "output_per_million": 0.800064
+ "input_per_million": 0.04,
+ "output_per_million": 0.04
}
}
},
"metadata": {
- "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).",
+ "description": "DeepSeek R1 Distill Llama 8B is a distilled large language model based on [Llama-3.1-8B-Instruct](/meta-llama/llama-3.1-8b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 50.4\n- MATH-500 pass@1: 89.1\n- CodeForces Rating: 1205\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.\n\nHugging Face: \n- [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) \n- [DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13360,44 +13785,40 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 32000,
+ "max_completion_tokens": 32000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-chat-v3-0324",
- "name": "DeepSeek: DeepSeek V3 0324",
+ "id": "deepseek/deepseek-r1-distill-qwen-1.5b",
+ "name": "DeepSeek: R1 Distill Qwen 1.5B",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-03-24 14:59:15 +0100",
- "context_window": 163840,
- "max_output_tokens": null,
+ "created_at": "2025-01-31 23:54:27 +1100",
+ "context_window": 131072,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13409,20 +13830,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1999188,
- "output_per_million": 0.800064
+ "input_per_million": 0.18,
+ "output_per_million": 0.18
}
}
},
"metadata": {
- "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.",
+ "description": "DeepSeek R1 Distill Qwen 1.5B is a distilled large language model based on [Qwen 2.5 Math 1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It's a very small and efficient model which outperforms [GPT 4o 0513](/openai/gpt-4o-2024-05-13) on Math Benchmarks.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 28.9\n- AIME 2024 cons@64: 52.7\n- MATH-500 pass@1: 83.9\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13431,44 +13850,39 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": null
+ "tokenizer": "Other",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 32768,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
- "response_format",
- "seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-chat-v3-0324:free",
- "name": "DeepSeek: DeepSeek V3 0324 (free)",
+ "id": "deepseek/deepseek-r1-distill-qwen-14b",
+ "name": "DeepSeek: R1 Distill Qwen 14B",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-03-24 14:59:15 +0100",
- "context_window": 163840,
- "max_output_tokens": 163840,
+ "created_at": "2025-01-30 10:39:00 +1100",
+ "context_window": 64000,
+ "max_output_tokens": 32000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13480,12 +13894,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.15,
+ "output_per_million": 0.15
+ }
+ }
+ },
"metadata": {
- "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.",
+ "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13494,41 +13914,39 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": 163840,
+ "context_length": 64000,
+ "max_completion_tokens": 32000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
"seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-chat-v3.1",
- "name": "DeepSeek: DeepSeek V3.1",
+ "id": "deepseek/deepseek-r1-distill-qwen-14b:free",
+ "name": "DeepSeek: R1 Distill Qwen 14B (free)",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-08-21 14:33:48 +0200",
- "context_window": 163840,
+ "created_at": "2025-01-30 10:39:00 +1100",
+ "context_window": 64000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -13541,20 +13959,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.7999999999999999
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference.\n\nThe model improves tool use, code generation, and reasoning efficiency, achieving performance comparable to DeepSeek-R1 on difficult benchmarks while responding more quickly. It supports structured tool calling, code agents, and search agents, making it suitable for research, coding, and agentic workflows.\n\nIt succeeds the [DeepSeek V3-0324](/deepseek/deepseek-chat-v3-0324) model and performs well on a variety of tasks.",
+ "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13563,11 +13972,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "deepseek-v3.1"
+ "tokenizer": "Qwen",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
+ "context_length": 64000,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -13582,13 +13991,9 @@
"presence_penalty",
"reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -13596,13 +14001,13 @@
}
},
{
- "id": "deepseek/deepseek-prover-v2",
- "name": "DeepSeek: DeepSeek Prover V2",
+ "id": "deepseek/deepseek-r1-distill-qwen-32b",
+ "name": "DeepSeek: R1 Distill Qwen 32B",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-04-30 13:38:14 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "created_at": "2025-01-30 10:53:50 +1100",
+ "context_window": 131072,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13614,18 +14019,19 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 2.1799999999999997
+ "input_per_million": 0.075,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "description": "DeepSeek Prover V2 is a 671B parameter model, speculated to be geared towards logic and mathematics. Likely an upgrade from [DeepSeek-Prover-V1.5](https://huggingface.co/deepseek-ai/DeepSeek-Prover-V1.5-RL) Not much is known about the model yet, as DeepSeek released it on Hugging Face without an announcement or description.",
+ "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\\n\\nOther benchmark results include:\\n\\n- AIME 2024 pass@1: 72.6\\n- MATH-500 pass@1: 94.3\\n- CodeForces Rating: 1691\\n\\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13634,20 +14040,23 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
"response_format",
"seed",
@@ -13659,13 +14068,13 @@
}
},
{
- "id": "deepseek/deepseek-r1",
- "name": "DeepSeek: R1",
+ "id": "deepseek/deepseek-r1:free",
+ "name": "DeepSeek: R1 (free)",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-01-20 14:51:35 +0100",
+ "created_at": "2025-01-21 00:51:35 +1100",
"context_window": 163840,
- "max_output_tokens": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13676,19 +14085,9 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "predicted_outputs"
+ "streaming"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 2.0
- }
- }
- },
+ "pricing": {},
"metadata": {
"description": "DeepSeek R1 is here: Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model & [technical report](https://api-docs.deepseek.com/news/news250120).\n\nMIT licensed: Distill & commercialize freely!",
"architecture": {
@@ -13704,39 +14103,24 @@
},
"top_provider": {
"context_length": 163840,
- "max_completion_tokens": 163840,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
- "response_format",
- "seed",
- "stop",
- "structured_outputs",
- "temperature",
- "tool_choice",
- "tools",
- "top_k",
- "top_logprobs",
- "top_p"
+ "temperature"
]
}
},
{
- "id": "deepseek/deepseek-r1-0528",
- "name": "DeepSeek: R1 0528",
+ "id": "deepseek/deepseek-v3.1-base",
+ "name": "DeepSeek: DeepSeek V3.1 Base",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-05-28 19:59:30 +0200",
+ "created_at": "2025-08-21 07:56:57 +1000",
"context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -13750,20 +14134,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1999188,
- "output_per_million": 0.800064
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.7999999999999999
}
}
},
"metadata": {
- "description": "May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model.",
+ "description": "This is a base model, trained only for raw next-token prediction. Unlike instruct/chat models, it has not been fine-tuned to follow user instructions. Prompts need to be written more like training text or examples rather than simple requests (e.g., “Translate the following sentence…” instead of just “Translate this”).\n\nDeepSeek-V3.1 Base is a 671B parameter open Mixture-of-Experts (MoE) language model with 37B active parameters per forward pass and a context length of 128K tokens. Trained on 14.8T tokens using FP8 mixed precision, it achieves high training efficiency and stability, with strong performance across language, reasoning, math, and coding tasks. \n",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13773,7 +14155,7 @@
"text"
],
"tokenizer": "DeepSeek",
- "instruct_type": "deepseek-r1"
+ "instruct_type": "none"
},
"top_provider": {
"context_length": 163840,
@@ -13783,21 +14165,15 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -13805,13 +14181,13 @@
}
},
{
- "id": "deepseek/deepseek-r1-0528-qwen3-8b",
- "name": "DeepSeek: Deepseek R1 0528 Qwen3 8B",
+ "id": "eleutherai/llemma_7b",
+ "name": "EleutherAI: Llemma 7b",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-05-29 19:09:03 +0200",
- "context_window": 32000,
- "max_output_tokens": null,
+ "family": "eleutherai",
+ "created_at": "2025-04-15 01:07:05 +1000",
+ "context_window": 4096,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13822,19 +14198,18 @@
]
},
"capabilities": [
- "streaming",
- "predicted_outputs"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01,
- "output_per_million": 0.02
+ "input_per_million": 0.7999999999999999,
+ "output_per_million": 1.2
}
}
},
"metadata": {
- "description": "DeepSeek-R1-0528 is a lightly upgraded release of DeepSeek R1 that taps more compute and smarter post-training tricks, pushing its reasoning and inference to the brink of flagship models like O3 and Gemini 2.5 Pro.\nIt now tops math, programming, and logic leaderboards, showcasing a step-change in depth-of-thought.\nThe distilled variant, DeepSeek-R1-0528-Qwen3-8B, transfers this chain-of-thought into an 8 B-parameter form, beating standard Qwen3 8B by +10 pp and tying the 235 B “thinking” giant on AIME 2024.",
+ "description": "Llemma 7B is a language model for mathematics. It was initialized with Code Llama 7B weights, and trained on the Proof-Pile-2 for 200B tokens. Llemma models are particularly strong at chain-of-thought mathematical reasoning and using computational tools for mathematics, such as Python and formal theorem provers.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13843,46 +14218,44 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Other",
+ "instruct_type": "code-llama"
},
"top_provider": {
- "context_length": 32000,
- "max_completion_tokens": null,
+ "context_length": 4096,
+ "max_completion_tokens": 4096,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
"seed",
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-0528-qwen3-8b:free",
- "name": "DeepSeek: Deepseek R1 0528 Qwen3 8B (free)",
+ "id": "google/gemini-2.0-flash-001",
+ "name": "Google: Gemini 2.0 Flash",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-05-29 19:09:03 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "family": "google",
+ "created_at": "2025-02-06 02:30:13 +1100",
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output": [
"text"
@@ -13890,59 +14263,66 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.39999999999999997,
+ "cached_input_per_million": 0.024999999999999998
+ }
+ }
+ },
"metadata": {
- "description": "DeepSeek-R1-0528 is a lightly upgraded release of DeepSeek R1 that taps more compute and smarter post-training tricks, pushing its reasoning and inference to the brink of flagship models like O3 and Gemini 2.5 Pro.\nIt now tops math, programming, and logic leaderboards, showcasing a step-change in depth-of-thought.\nThe distilled variant, DeepSeek-R1-0528-Qwen3-8B, transfers this chain-of-thought into an 8 B-parameter form, beating standard Qwen3 8B by +10 pp and tying the 235 B “thinking” giant on AIME 2024.",
+ "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
- "top_logprobs",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-0528:free",
- "name": "DeepSeek: R1 0528 (free)",
+ "id": "google/gemini-2.0-flash-exp:free",
+ "name": "Google: Gemini 2.0 Flash Experimental (free)",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-05-28 19:59:30 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "google",
+ "created_at": "2024-12-12 04:18:43 +1100",
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -13950,59 +14330,57 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {},
"metadata": {
- "description": "May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model.",
+ "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
"temperature",
- "top_k",
- "top_logprobs",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-llama-70b",
- "name": "DeepSeek: R1 Distill Llama 70B",
+ "id": "google/gemini-2.0-flash-lite-001",
+ "name": "Google: Gemini 2.0 Flash Lite",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-23 21:12:49 +0100",
- "context_window": 131072,
- "max_output_tokens": null,
+ "family": "google",
+ "created_at": "2025-02-26 04:56:52 +1100",
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output": [
"text"
@@ -14011,70 +14389,66 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.0333198,
- "output_per_million": 0.13334400000000002
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "Gemini 2.0 Flash Lite offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5), all at extremely economical token prices.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-llama-70b:free",
- "name": "DeepSeek: R1 Distill Llama 70B (free)",
+ "id": "google/gemini-2.5-flash",
+ "name": "Google: Gemini 2.5 Flash",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-23 21:12:49 +0100",
- "context_window": 8192,
- "max_output_tokens": 4096,
+ "family": "google",
+ "created_at": "2025-06-18 01:01:28 +1000",
+ "context_window": 1048576,
+ "max_output_tokens": 65535,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output": [
"text"
@@ -14082,188 +14456,187 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.3,
+ "output_per_million": 2.5,
+ "cached_input_per_million": 0.075
+ }
+ }
+ },
"metadata": {
- "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning).",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 4096,
+ "context_length": 1048576,
+ "max_completion_tokens": 65535,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
- "top_logprobs",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-llama-8b",
- "name": "DeepSeek: R1 Distill Llama 8B",
+ "id": "google/gemini-2.5-flash-image-preview",
+ "name": "Google: Gemini 2.5 Flash Image Preview",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-02-07 15:15:18 +0100",
- "context_window": 32000,
- "max_output_tokens": 32000,
+ "family": "google",
+ "created_at": "2025-08-27 00:36:17 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "image",
"text"
]
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.04,
- "output_per_million": 0.04
+ "input_per_million": 0.3,
+ "output_per_million": 2.5
}
}
},
"metadata": {
- "description": "DeepSeek R1 Distill Llama 8B is a distilled large language model based on [Llama-3.1-8B-Instruct](/meta-llama/llama-3.1-8b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 50.4\n- MATH-500 pass@1: 89.1\n- CodeForces Rating: 1205\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.\n\nHugging Face: \n- [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) \n- [DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |",
+ "description": "Gemini 2.5 Flash Image Preview is a state of the art image generation model with contextual understanding. It is capable of image generation, edits, and multi-turn conversations.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text+image",
"input_modalities": [
+ "image",
"text"
],
"output_modalities": [
+ "image",
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32000,
- "max_completion_tokens": 32000,
+ "context_length": 32768,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
"max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
- "stop",
+ "structured_outputs",
"temperature",
- "top_k",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-qwen-1.5b",
- "name": "DeepSeek: R1 Distill Qwen 1.5B",
+ "id": "google/gemini-2.5-flash-image-preview:free",
+ "name": "Google: Gemini 2.5 Flash Image Preview (free)",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-31 13:54:27 +0100",
- "context_window": 131072,
- "max_output_tokens": 32768,
+ "family": "google",
+ "created_at": "2025-08-27 00:36:17 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "image",
"text"
]
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.18,
- "output_per_million": 0.18
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "DeepSeek R1 Distill Qwen 1.5B is a distilled large language model based on [Qwen 2.5 Math 1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It's a very small and efficient model which outperforms [GPT 4o 0513](/openai/gpt-4o-2024-05-13) on Math Benchmarks.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 28.9\n- AIME 2024 cons@64: 52.7\n- MATH-500 pass@1: 83.9\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "Gemini 2.5 Flash Image Preview is a state of the art image generation model with contextual understanding. It is capable of image generation, edits, and multi-turn conversations.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text+image",
"input_modalities": [
+ "image",
"text"
],
"output_modalities": [
+ "image",
"text"
],
- "tokenizer": "Other",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 32768,
+ "context_length": 32768,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
"max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
- "stop",
+ "response_format",
+ "seed",
+ "structured_outputs",
"temperature",
- "top_k",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-qwen-14b",
- "name": "DeepSeek: R1 Distill Qwen 14B",
+ "id": "google/gemini-2.5-flash-lite",
+ "name": "Google: Gemini 2.5 Flash Lite",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-30 00:39:00 +0100",
- "context_window": 64000,
- "max_output_tokens": 32000,
+ "family": "google",
+ "created_at": "2025-07-23 02:04:36 +1000",
+ "context_window": 1048576,
+ "max_output_tokens": 65535,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output": [
"text"
@@ -14271,64 +14644,70 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.15
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.39999999999999997,
+ "cached_input_per_million": 0.024999999999999998
}
}
},
"metadata": {
- "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, \"thinking\" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. ",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 64000,
- "max_completion_tokens": 32000,
+ "context_length": 1048576,
+ "max_completion_tokens": 65535,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-qwen-14b:free",
- "name": "DeepSeek: R1 Distill Qwen 14B (free)",
+ "id": "google/gemini-2.5-flash-lite-preview-06-17",
+ "name": "Google: Gemini 2.5 Flash Lite Preview 06-17",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-30 00:39:00 +0100",
- "context_window": 64000,
- "max_output_tokens": null,
+ "family": "google",
+ "created_at": "2025-06-18 01:23:51 +1000",
+ "context_window": 1048576,
+ "max_output_tokens": 65535,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output": [
"text"
@@ -14336,59 +14715,70 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.39999999999999997,
+ "cached_input_per_million": 0.024999999999999998
+ }
+ }
+ },
"metadata": {
- "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, \"thinking\" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. ",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 64000,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 65535,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
- "top_logprobs",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-qwen-32b",
- "name": "DeepSeek: R1 Distill Qwen 32B",
+ "id": "google/gemini-2.5-pro",
+ "name": "Google: Gemini 2.5 Pro",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-30 00:53:50 +0100",
- "context_window": 131072,
- "max_output_tokens": 16384,
+ "family": "google",
+ "created_at": "2025-06-18 00:12:24 +1000",
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output": [
"text"
@@ -14396,114 +14786,129 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.15
+ "input_per_million": 1.25,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 0.31
}
}
},
"metadata": {
- "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\\n\\nOther benchmark results include:\\n\\n- AIME 2024 pass@1: 72.6\\n- MATH-500 pass@1: 94.3\\n- CodeForces Rating: 1691\\n\\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 16384,
+ "context_length": 1048576,
+ "max_completion_tokens": 65536,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1:free",
- "name": "DeepSeek: R1 (free)",
+ "id": "google/gemini-2.5-pro-exp-03-25",
+ "name": "Google: Gemini 2.5 Pro Experimental",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-20 14:51:35 +0100",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "google",
+ "created_at": "2025-03-26 04:01:39 +1100",
+ "context_window": 1048576,
+ "max_output_tokens": 65535,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {},
"metadata": {
- "description": "DeepSeek R1 is here: Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model & [technical report](https://api-docs.deepseek.com/news/news250120).\n\nMIT licensed: Distill & commercialize freely!",
+ "description": "This model has been deprecated by Google in favor of the (paid Preview model)[google/gemini-2.5-pro-preview]\n \nGemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
- ],
- "tokenizer": "DeepSeek",
- "instruct_type": "deepseek-r1"
+ ],
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 65535,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
"max_tokens",
- "reasoning",
- "temperature"
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
+ "temperature",
+ "tool_choice",
+ "tools",
+ "top_p"
]
}
},
{
- "id": "deepseek/deepseek-v3-base",
- "name": "DeepSeek: DeepSeek V3 Base",
+ "id": "google/gemini-2.5-pro-preview",
+ "name": "Google: Gemini 2.5 Pro Preview 06-05",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-03-29 19:13:43 +0100",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "google",
+ "created_at": "2025-06-06 01:27:37 +1000",
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output": [
"text"
@@ -14511,64 +14916,70 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1999188,
- "output_per_million": 0.800064
+ "input_per_million": 1.25,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 0.31
}
}
},
"metadata": {
- "description": "Note that this is a base model mostly meant for testing, you need to provide detailed prompts for the model to return useful responses. \n\nDeepSeek-V3 Base is a 671B parameter open Mixture-of-Experts (MoE) language model with 37B active parameters per forward pass and a context length of 128K tokens. Trained on 14.8T tokens using FP8 mixed precision, it achieves high training efficiency and stability, with strong performance across language, reasoning, math, and coding tasks. \n\nDeepSeek-V3 Base is the pre-trained model behind [DeepSeek V3](/deepseek/deepseek-chat-v3)",
+ "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.\n",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
+ "tokenizer": "Gemini",
"instruct_type": null
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 65536,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
+ "include_reasoning",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
+ "reasoning",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
- "top_logprobs",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-v3.1-base",
- "name": "DeepSeek: DeepSeek V3.1 Base",
+ "id": "google/gemini-2.5-pro-preview-05-06",
+ "name": "Google: Gemini 2.5 Pro Preview 05-06",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-08-20 23:56:57 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "google",
+ "created_at": "2025-05-07 10:41:53 +1000",
+ "context_window": 1048576,
+ "max_output_tokens": 65535,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output": [
"text"
@@ -14576,182 +14987,202 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.7999999999999999
+ "input_per_million": 1.25,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 0.31
}
}
},
"metadata": {
- "description": "This is a base model, trained only for raw next-token prediction. Unlike instruct/chat models, it has not been fine-tuned to follow user instructions. Prompts need to be written more like training text or examples rather than simple requests (e.g., “Translate the following sentence…” instead of just “Translate this”).\n\nDeepSeek-V3.1 Base is a 671B parameter open Mixture-of-Experts (MoE) language model with 37B active parameters per forward pass and a context length of 128K tokens. Trained on 14.8T tokens using FP8 mixed precision, it achieves high training efficiency and stability, with strong performance across language, reasoning, math, and coding tasks. \n",
+ "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "none"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 65535,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
+ "include_reasoning",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
+ "reasoning",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
- "top_logprobs",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "eleutherai/llemma_7b",
- "name": "EleutherAI: Llemma 7b",
+ "id": "google/gemini-flash-1.5",
+ "name": "Google: Gemini 1.5 Flash ",
"provider": "openrouter",
- "family": "eleutherai",
- "created_at": "2025-04-14 17:07:05 +0200",
- "context_window": 4096,
- "max_output_tokens": 4096,
+ "family": "google",
+ "created_at": "2024-05-14 10:00:00 +1000",
+ "context_window": 1000000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7999999999999999,
- "output_per_million": 1.2
+ "input_per_million": 0.075,
+ "output_per_million": 0.3,
+ "cached_input_per_million": 0.01875
}
}
},
"metadata": {
- "description": "Llemma 7B is a language model for mathematics. It was initialized with Code Llama 7B weights, and trained on the Proof-Pile-2 for 200B tokens. Llemma models are particularly strong at chain-of-thought mathematical reasoning and using computational tools for mathematics, such as Python and formal theorem provers.",
+ "description": "Gemini 1.5 Flash is a foundation model that performs well at a variety of multimodal tasks such as visual understanding, classification, summarization, and creating content from image, audio and video. It's adept at processing visual and text inputs such as photographs, documents, infographics, and screenshots.\n\nGemini 1.5 Flash is designed for high-volume, high-frequency tasks where cost and latency matter. On most common tasks, Flash achieves comparable quality to other Gemini Pro models at a significantly reduced cost. Flash is well-suited for applications like chat assistants and on-demand content generation where speed and scale matter.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "code-llama"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 4096,
- "max_completion_tokens": 4096,
+ "context_length": 1000000,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "featherless/qwerky-72b:free",
- "name": "Qrwkv 72B (free)",
+ "id": "google/gemini-flash-1.5-8b",
+ "name": "Google: Gemini 1.5 Flash 8B",
"provider": "openrouter",
- "family": "featherless",
- "created_at": "2025-03-20 15:39:57 +0100",
- "context_window": 32768,
- "max_output_tokens": 4096,
+ "family": "google",
+ "created_at": "2024-10-03 10:00:00 +1000",
+ "context_window": 1000000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15,
+ "cached_input_per_million": 0.01
+ }
+ }
+ },
"metadata": {
- "description": "Qrwkv-72B is a linear-attention RWKV variant of the Qwen 2.5 72B model, optimized to significantly reduce computational cost at scale. Leveraging linear attention, it achieves substantial inference speedups (>1000x) while retaining competitive accuracy on common benchmarks like ARC, HellaSwag, Lambada, and MMLU. It inherits knowledge and language support from Qwen 2.5, supporting approximately 30 languages, making it suitable for efficient inference in large-context applications.",
+ "description": "Gemini Flash 1.5 8B is optimized for speed and efficiency, offering enhanced performance in small prompt tasks like chat, transcription, and translation. With reduced latency, it is highly effective for real-time and large-scale operations. This model focuses on cost-effective solutions while maintaining high-quality results.\n\n[Click here to learn more about this model](https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Gemini",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 4096,
+ "context_length": 1000000,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "google/gemini-2.0-flash-001",
- "name": "Google: Gemini 2.0 Flash",
+ "id": "google/gemini-pro-1.5",
+ "name": "Google: Gemini 1.5 Pro",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-02-05 16:30:13 +0100",
- "context_window": 1048576,
+ "created_at": "2024-04-09 10:00:00 +1000",
+ "context_window": 2000000,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text",
- "image",
- "file",
- "audio"
+ "image"
],
"output": [
"text"
@@ -14765,21 +15196,18 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.39999999999999997,
- "cached_input_per_million": 0.024999999999999998
+ "input_per_million": 1.25,
+ "output_per_million": 5.0
}
}
},
"metadata": {
- "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.",
+ "description": "Google's latest multimodal model, supports image and video[0] in text or chat prompts.\n\nOptimized for language tasks including:\n\n- Code generation\n- Text generation\n- Text editing\n- Problem solving\n- Recommendations\n- Information extraction\n- Data extraction or generation\n- AI agents\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n* [0]: Video input is not available through OpenRouter at this time.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
"text",
- "image",
- "file",
- "audio"
+ "image"
],
"output_modalities": [
"text"
@@ -14788,13 +15216,15 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 1048576,
+ "context_length": 2000000,
"max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"max_tokens",
+ "presence_penalty",
"response_format",
"seed",
"stop",
@@ -14807,76 +15237,82 @@
}
},
{
- "id": "google/gemini-2.0-flash-exp:free",
- "name": "Google: Gemini 2.0 Flash Experimental (free)",
+ "id": "google/gemma-2-27b-it",
+ "name": "Google: Gemma 2 27B",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-12-11 18:18:43 +0100",
- "context_window": 1048576,
- "max_output_tokens": 8192,
+ "created_at": "2024-07-13 10:00:00 +1000",
+ "context_window": 8192,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
]
},
- "capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
- ],
- "pricing": {},
+ "capabilities": [
+ "streaming",
+ "structured_output",
+ "predicted_outputs"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.65,
+ "output_per_million": 0.65
+ }
+ }
+ },
"metadata": {
- "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.",
+ "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 8192,
+ "context_length": 8192,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"response_format",
- "seed",
"stop",
+ "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
"top_p"
]
}
},
{
- "id": "google/gemini-2.0-flash-lite-001",
- "name": "Google: Gemini 2.0 Flash Lite",
+ "id": "google/gemma-2-9b-it",
+ "name": "Google: Gemma 2 9B",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-02-25 18:56:52 +0100",
- "context_window": 1048576,
+ "created_at": "2024-06-28 10:00:00 +1000",
+ "context_window": 8192,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file",
- "audio"
+ "text"
],
"output": [
"text"
@@ -14884,67 +15320,66 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
+ "input_per_million": 0.01,
+ "output_per_million": 0.0100008
}
}
},
"metadata": {
- "description": "Gemini 2.0 Flash Lite offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5), all at extremely economical token prices.",
+ "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file",
- "audio"
+ "text"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
+ "context_length": 8192,
"max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemini-2.5-flash",
- "name": "Google: Gemini 2.5 Flash",
+ "id": "google/gemma-2-9b-it:free",
+ "name": "Google: Gemma 2 9B (free)",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-06-17 17:01:28 +0200",
- "context_window": 1048576,
- "max_output_tokens": 65535,
+ "created_at": "2024-06-28 10:00:00 +1000",
+ "context_window": 8192,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "file",
- "image",
- "text",
- "audio"
+ "text"
],
"output": [
"text"
@@ -14952,70 +15387,58 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.3,
- "output_per_million": 2.5,
- "cached_input_per_million": 0.075
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning).",
+ "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "file",
- "image",
- "text",
- "audio"
+ "text"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 65535,
+ "context_length": 8192,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
- "reasoning",
- "response_format",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemini-2.5-flash-lite",
- "name": "Google: Gemini 2.5 Flash Lite",
+ "id": "google/gemma-3-12b-it",
+ "name": "Google: Gemma 3 12B",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-07-22 18:04:36 +0200",
- "context_window": 1048576,
- "max_output_tokens": 65535,
+ "created_at": "2025-03-14 08:50:25 +1100",
+ "context_window": 96000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output": [
"text"
@@ -15023,70 +15446,68 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.39999999999999997,
- "cached_input_per_million": 0.024999999999999998
+ "input_per_million": 0.0481286,
+ "output_per_million": 0.192608
}
}
},
"metadata": {
- "description": "Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, \"thinking\" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. ",
+ "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 65535,
+ "context_length": 96000,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
- "reasoning",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemini-2.5-flash-lite-preview-06-17",
- "name": "Google: Gemini 2.5 Flash Lite Preview 06-17",
+ "id": "google/gemma-3-12b-it:free",
+ "name": "Google: Gemma 3 12B (free)",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-06-17 17:23:51 +0200",
- "context_window": 1048576,
- "max_output_tokens": 65535,
+ "created_at": "2025-03-14 08:50:25 +1100",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output": [
"text"
@@ -15094,70 +15515,59 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.39999999999999997,
- "cached_input_per_million": 0.024999999999999998
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, \"thinking\" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. ",
+ "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 65535,
+ "context_length": 32768,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
- "reasoning",
- "response_format",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemini-2.5-pro",
- "name": "Google: Gemini 2.5 Pro",
+ "id": "google/gemma-3-27b-it",
+ "name": "Google: Gemma 3 27B",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-06-17 16:12:24 +0200",
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "created_at": "2025-03-12 16:12:39 +1100",
+ "context_window": 96000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output": [
"text"
@@ -15165,69 +15575,69 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.31
+ "input_per_million": 0.0666396,
+ "output_per_million": 0.26668800000000004
}
}
},
"metadata": {
- "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.",
+ "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 65536,
+ "context_length": 96000,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
- "reasoning",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemini-2.5-pro-exp-03-25",
- "name": "Google: Gemini 2.5 Pro Experimental",
+ "id": "google/gemma-3-27b-it:free",
+ "name": "Google: Gemma 3 27B (free)",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-03-25 18:01:39 +0100",
- "context_window": 1048576,
- "max_output_tokens": 65535,
+ "created_at": "2025-03-12 16:12:39 +1100",
+ "context_window": 96000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text",
- "image",
- "file"
+ "image"
],
"output": [
"text"
@@ -15235,59 +15645,62 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {},
"metadata": {
- "description": "This model has been deprecated by Google in favor of the (paid Preview model)[google/gemini-2.5-pro-preview]\n \nGemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.",
+ "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
"text",
- "image",
- "file"
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 65535,
+ "context_length": 96000,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemini-2.5-pro-preview",
- "name": "Google: Gemini 2.5 Pro Preview 06-05",
+ "id": "google/gemma-3-4b-it",
+ "name": "Google: Gemma 3 4B",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-06-05 17:27:37 +0200",
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "created_at": "2025-03-14 09:38:30 +1100",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output": [
"text"
@@ -15295,70 +15708,64 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.31
+ "input_per_million": 0.02,
+ "output_per_million": 0.04
}
}
},
"metadata": {
- "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.\n",
+ "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 65536,
+ "context_length": 131072,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
"max_tokens",
- "reasoning",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
"top_p"
]
}
},
{
- "id": "google/gemini-2.5-pro-preview-05-06",
- "name": "Google: Gemini 2.5 Pro Preview 05-06",
+ "id": "google/gemma-3-4b-it:free",
+ "name": "Google: Gemma 3 4B (free)",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-05-07 02:41:53 +0200",
- "context_window": 1048576,
- "max_output_tokens": 65535,
+ "created_at": "2025-03-14 09:38:30 +1100",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text",
- "image",
- "file",
- "audio"
+ "image"
],
"output": [
"text"
@@ -15366,68 +15773,51 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.31
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.",
+ "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
"text",
- "image",
- "file",
- "audio"
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 65535,
+ "context_length": 32768,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
"max_tokens",
- "reasoning",
"response_format",
"seed",
- "stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_p"
]
}
},
{
- "id": "google/gemini-flash-1.5",
- "name": "Google: Gemini 1.5 Flash ",
+ "id": "google/gemma-3n-e2b-it:free",
+ "name": "Google: Gemma 3n 2B (free)",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-05-14 02:00:00 +0200",
- "context_window": 1000000,
- "max_output_tokens": 8192,
+ "created_at": "2025-07-10 01:28:24 +1000",
+ "context_window": 8192,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -15435,35 +15825,25 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3,
- "cached_input_per_million": 0.01875
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Gemini 1.5 Flash is a foundation model that performs well at a variety of multimodal tasks such as visual understanding, classification, summarization, and creating content from image, audio and video. It's adept at processing visual and text inputs such as photographs, documents, infographics, and screenshots.\n\nGemini 1.5 Flash is designed for high-volume, high-frequency tasks where cost and latency matter. On most common tasks, Flash achieves comparable quality to other Gemini Pro models at a significantly reduced cost. Flash is well-suited for applications like chat assistants and on-demand content generation where speed and scale matter.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal",
+ "description": "Gemma 3n E2B IT is a multimodal, instruction-tuned model developed by Google DeepMind, designed to operate efficiently at an effective parameter size of 2B while leveraging a 6B architecture. Based on the MatFormer architecture, it supports nested submodels and modular composition via the Mix-and-Match framework. Gemma 3n models are optimized for low-resource deployment, offering 32K context length and strong multilingual and reasoning performance across common benchmarks. This variant is trained on a diverse corpus including code, math, web, and multimodal data.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 1000000,
- "max_completion_tokens": 8192,
+ "context_length": 8192,
+ "max_completion_tokens": 2048,
"is_moderated": false
},
"per_request_limits": null,
@@ -15474,27 +15854,23 @@
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_p"
]
}
},
{
- "id": "google/gemini-flash-1.5-8b",
- "name": "Google: Gemini 1.5 Flash 8B",
+ "id": "google/gemma-3n-e4b-it",
+ "name": "Google: Gemma 3n 4B",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-10-03 02:00:00 +0200",
- "context_window": 1000000,
- "max_output_tokens": 8192,
+ "created_at": "2025-05-21 07:33:44 +1000",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -15502,66 +15878,61 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15,
- "cached_input_per_million": 0.01
+ "input_per_million": 0.02,
+ "output_per_million": 0.04
}
}
},
"metadata": {
- "description": "Gemini Flash 1.5 8B is optimized for speed and efficiency, offering enhanced performance in small prompt tasks like chat, transcription, and translation. With reduced latency, it is highly effective for real-time and large-scale operations. This model focuses on cost-effective solutions while maintaining high-quality results.\n\n[Click here to learn more about this model](https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).",
+ "description": "Gemma 3n E4B-it is optimized for efficient execution on mobile and low-resource devices, such as phones, laptops, and tablets. It supports multimodal inputs—including text, visual data, and audio—enabling diverse tasks such as text generation, speech recognition, translation, and image analysis. Leveraging innovations like Per-Layer Embedding (PLE) caching and the MatFormer architecture, Gemma 3n dynamically manages memory usage and computational load by selectively activating model parameters, significantly reducing runtime resource requirements.\n\nThis model supports a wide linguistic range (trained in over 140 languages) and features a flexible 32K token context window. Gemma 3n can selectively load parameters, optimizing memory and computational efficiency based on the task or device capabilities, making it well-suited for privacy-focused, offline-capable applications and on-device AI solutions. [Read more in the blog post](https://developers.googleblog.com/en/introducing-gemma-3n/)",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 1000000,
- "max_completion_tokens": 8192,
+ "context_length": 32768,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
- "seed",
+ "repetition_penalty",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
"top_p"
]
}
},
{
- "id": "google/gemini-pro-1.5",
- "name": "Google: Gemini 1.5 Pro",
+ "id": "google/gemma-3n-e4b-it:free",
+ "name": "Google: Gemma 3n 4B (free)",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-04-09 02:00:00 +0200",
- "context_window": 2000000,
- "max_output_tokens": 8192,
+ "created_at": "2025-05-21 07:33:44 +1000",
+ "context_window": 8192,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -15569,34 +15940,25 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 1.25,
- "output_per_million": 5.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Google's latest multimodal model, supports image and video[0] in text or chat prompts.\n\nOptimized for language tasks including:\n\n- Code generation\n- Text generation\n- Text editing\n- Problem solving\n- Recommendations\n- Information extraction\n- Data extraction or generation\n- AI agents\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n* [0]: Video input is not available through OpenRouter at this time.",
+ "description": "Gemma 3n E4B-it is optimized for efficient execution on mobile and low-resource devices, such as phones, laptops, and tablets. It supports multimodal inputs—including text, visual data, and audio—enabling diverse tasks such as text generation, speech recognition, translation, and image analysis. Leveraging innovations like Per-Layer Embedding (PLE) caching and the MatFormer architecture, Gemma 3n dynamically manages memory usage and computational load by selectively activating model parameters, significantly reducing runtime resource requirements.\n\nThis model supports a wide linguistic range (trained in over 140 languages) and features a flexible 32K token context window. Gemma 3n can selectively load parameters, optimizing memory and computational efficiency based on the task or device capabilities, making it well-suited for privacy-focused, offline-capable applications and on-device AI solutions. [Read more in the blog post](https://developers.googleblog.com/en/introducing-gemma-3n/)",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 2000000,
- "max_completion_tokens": 8192,
+ "context_length": 8192,
+ "max_completion_tokens": 2048,
"is_moderated": false
},
"per_request_limits": null,
@@ -15607,21 +15969,18 @@
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_p"
]
}
},
{
- "id": "google/gemma-2-27b-it",
- "name": "Google: Gemma 2 27B",
+ "id": "gryphe/mythomax-l2-13b",
+ "name": "MythoMax 13B",
"provider": "openrouter",
- "family": "google",
- "created_at": "2024-07-13 02:00:00 +0200",
- "context_window": 8192,
+ "family": "gryphe",
+ "created_at": "2023-07-02 10:00:00 +1000",
+ "context_window": 4096,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -15640,13 +15999,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.65,
- "output_per_million": 0.65
+ "input_per_million": 0.06,
+ "output_per_million": 0.06
}
}
},
"metadata": {
- "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
+ "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -15655,11 +16014,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
+ "tokenizer": "Llama2",
+ "instruct_type": "alpaca"
},
"top_provider": {
- "context_length": 8192,
+ "context_length": 4096,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -15672,22 +16031,24 @@
"presence_penalty",
"repetition_penalty",
"response_format",
+ "seed",
"stop",
"structured_outputs",
"temperature",
+ "top_a",
"top_k",
"top_p"
]
}
},
{
- "id": "google/gemma-2-9b-it",
- "name": "Google: Gemma 2 9B",
+ "id": "inception/mercury",
+ "name": "Inception: Mercury",
"provider": "openrouter",
- "family": "google",
- "created_at": "2024-06-28 02:00:00 +0200",
- "context_window": 8192,
- "max_output_tokens": 8192,
+ "family": "inception",
+ "created_at": "2025-06-27 07:23:46 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -15699,19 +16060,19 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01,
- "output_per_million": 0.0100008
+ "input_per_million": 0.25,
+ "output_per_million": 1.0
}
}
},
"metadata": {
- "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
+ "description": "Mercury is the first diffusion large language model (dLLM). Applying a breakthrough discrete diffusion approach, the model runs 5-10x faster than even speed optimized models like GPT-4.1 Nano and Claude 3.5 Haiku while matching their performance. Mercury's speed enables developers to provide responsive user experiences, including with voice agents, search interfaces, and chatbots. Read more in the blog post here. ",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -15720,41 +16081,38 @@
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 8192,
+ "context_length": 128000,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
- "seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-2-9b-it:free",
- "name": "Google: Gemma 2 9B (free)",
+ "id": "inception/mercury-coder",
+ "name": "Inception: Mercury Coder",
"provider": "openrouter",
- "family": "google",
- "created_at": "2024-06-28 02:00:00 +0200",
- "context_window": 8192,
- "max_output_tokens": 8192,
+ "family": "inception",
+ "created_at": "2025-05-01 03:24:40 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -15766,11 +16124,19 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.25,
+ "output_per_million": 1.0
+ }
+ }
+ },
"metadata": {
- "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
+ "description": "Mercury Coder is the first diffusion large language model (dLLM). Applying a breakthrough discrete diffusion approach, the model runs 5-10x faster than even speed optimized models like Claude 3.5 Haiku and GPT-4o Mini while matching their performance. Mercury Coder's speed means that developers can stay in the flow while coding, enjoying rapid chat-based iteration and responsive code completion suggestions. On Copilot Arena, Mercury Coder ranks 1st in speed and ties for 2nd in quality. Read more in the [blog post here](https://www.inceptionlabs.ai/introducing-mercury).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -15779,307 +16145,274 @@
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 8192,
+ "context_length": 128000,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
- "seed",
+ "response_format",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-3-12b-it",
- "name": "Google: Gemma 3 12B",
+ "id": "infermatic/mn-inferor-12b",
+ "name": "Infermatic: Mistral Nemo Inferor 12B",
"provider": "openrouter",
- "family": "google",
- "created_at": "2025-03-13 22:50:25 +0100",
- "context_window": 96000,
+ "family": "infermatic",
+ "created_at": "2024-11-13 13:20:28 +1100",
+ "context_window": 8192,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "structured_output",
- "predicted_outputs"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.0481286,
- "output_per_million": 0.192608
+ "input_per_million": 0.6,
+ "output_per_million": 1.0
}
}
},
"metadata": {
- "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)",
+ "description": "Inferor 12B is a merge of top roleplay models, expert on immersive narratives and storytelling.\n\nThis model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [anthracite-org/magnum-v4-12b](https://openrouter.ai/anthracite-org/magnum-v4-72b) as a base.\n",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
+ "tokenizer": "Mistral",
+ "instruct_type": "mistral"
},
"top_provider": {
- "context_length": 96000,
+ "context_length": 8192,
"max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
- "response_format",
"seed",
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-3-12b-it:free",
- "name": "Google: Gemma 3 12B (free)",
+ "id": "inflection/inflection-3-pi",
+ "name": "Inflection: Inflection 3 Pi",
"provider": "openrouter",
- "family": "google",
- "created_at": "2025-03-13 22:50:25 +0100",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "family": "inflection",
+ "created_at": "2024-10-11 11:00:00 +1100",
+ "context_window": 8000,
+ "max_output_tokens": 1024,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "predicted_outputs"
+ "streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
+ }
+ }
+ },
"metadata": {
- "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)",
+ "description": "Inflection 3 Pi powers Inflection's [Pi](https://pi.ai) chatbot, including backstory, emotional intelligence, productivity, and safety. It has access to recent news, and excels in scenarios like customer support and roleplay.\n\nPi has been trained to mirror your tone and style, if you use more emojis, so will Pi! Try experimenting with various prompts and conversation styles.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 8192,
+ "context_length": 8000,
+ "max_completion_tokens": 1024,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "seed",
"stop",
"temperature",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-3-27b-it",
- "name": "Google: Gemma 3 27B",
+ "id": "inflection/inflection-3-productivity",
+ "name": "Inflection: Inflection 3 Productivity",
"provider": "openrouter",
- "family": "google",
- "created_at": "2025-03-12 06:12:39 +0100",
- "context_window": 96000,
- "max_output_tokens": 8192,
+ "family": "inflection",
+ "created_at": "2024-10-11 11:00:00 +1100",
+ "context_window": 8000,
+ "max_output_tokens": 1024,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "structured_output",
- "predicted_outputs"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.0666396,
- "output_per_million": 0.26668800000000004
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)",
+ "description": "Inflection 3 Productivity is optimized for following instructions. It is better for tasks requiring JSON output or precise adherence to provided guidelines. It has access to recent news.\n\nFor emotional intelligence similar to Pi, see [Inflect 3 Pi](/inflection/inflection-3-pi)\n\nSee [Inflection's announcement](https://inflection.ai/blog/enterprise) for more details.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 96000,
- "max_completion_tokens": 8192,
+ "context_length": 8000,
+ "max_completion_tokens": 1024,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "response_format",
- "seed",
"stop",
- "structured_outputs",
"temperature",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-3-27b-it:free",
- "name": "Google: Gemma 3 27B (free)",
+ "id": "liquid/lfm-3b",
+ "name": "Liquid: LFM 3B",
"provider": "openrouter",
- "family": "google",
- "created_at": "2025-03-12 06:12:39 +0100",
- "context_window": 96000,
- "max_output_tokens": 8192,
+ "family": "liquid",
+ "created_at": "2025-01-25 23:01:41 +1100",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "structured_output",
- "predicted_outputs"
+ "streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.02,
+ "output_per_million": 0.02
+ }
+ }
+ },
"metadata": {
- "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)",
+ "description": "Liquid's LFM 3B delivers incredible performance for its size. It positions itself as first place among 3B parameter transformers, hybrids, and RNN models It is also on par with Phi-3.5-mini on multiple benchmarks, while being 18.4% smaller.\n\nLFM-3B is the ideal choice for mobile and other edge text-based applications.\n\nSee the [launch announcement](https://www.liquid.ai/liquid-foundation-models) for benchmarks and more info.",
"architecture": {
- "modality": "text+image->text",
- "input_modalities": [
- "text",
- "image"
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
+ "tokenizer": "Other",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 96000,
- "max_completion_tokens": 8192,
+ "context_length": 32768,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-3-4b-it",
- "name": "Google: Gemma 3 4B",
+ "id": "liquid/lfm-7b",
+ "name": "Liquid: LFM 7B",
"provider": "openrouter",
- "family": "google",
- "created_at": "2025-03-13 23:38:30 +0100",
- "context_window": 131072,
+ "family": "liquid",
+ "created_at": "2025-01-25 23:08:03 +1100",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -16087,38 +16420,40 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.02,
- "output_per_million": 0.04
+ "input_per_million": 0.01,
+ "output_per_million": 0.01
}
}
},
"metadata": {
- "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.",
+ "description": "LFM-7B, a new best-in-class language model. LFM-7B is designed for exceptional chat capabilities, including languages like Arabic and Japanese. Powered by the Liquid Foundation Model (LFM) architecture, it exhibits unique features like low memory footprint and fast inference speed. \n\nLFM-7B is the world’s best-in-class multilingual language model in English, Arabic, and Japanese.\n\nSee the [launch announcement](https://www.liquid.ai/lfm-7b) for benchmarks and more info.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
+ "tokenizer": "Other",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -16128,23 +16463,23 @@
"stop",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-3-4b-it:free",
- "name": "Google: Gemma 3 4B (free)",
+ "id": "mancer/weaver",
+ "name": "Mancer: Weaver (alpha)",
"provider": "openrouter",
- "family": "google",
- "created_at": "2025-03-13 23:38:30 +0100",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "family": "mancer",
+ "created_at": "2023-08-02 10:00:00 +1000",
+ "context_window": 8000,
+ "max_output_tokens": 2000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -16152,47 +16487,59 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.125,
+ "output_per_million": 1.125
+ }
+ }
+ },
"metadata": {
- "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.",
+ "description": "An attempt to recreate Claude-style verbosity, but don't expect the same level of coherence or memory. Meant for use in roleplay/narrative situations.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
+ "tokenizer": "Llama2",
+ "instruct_type": "alpaca"
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 8192,
+ "context_length": 8000,
+ "max_completion_tokens": 2000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
- "response_format",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"seed",
- "structured_outputs",
+ "stop",
"temperature",
+ "top_a",
+ "top_k",
"top_p"
]
}
},
{
- "id": "google/gemma-3n-e2b-it:free",
- "name": "Google: Gemma 3n 2B (free)",
+ "id": "meta-llama/llama-3-70b-instruct",
+ "name": "Meta: Llama 3 70B Instruct",
"provider": "openrouter",
- "family": "google",
- "created_at": "2025-07-09 17:28:24 +0200",
+ "family": "meta-llama",
+ "created_at": "2024-04-18 10:00:00 +1000",
"context_window": 8192,
- "max_output_tokens": 2048,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -16204,11 +16551,20 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.3,
+ "output_per_million": 0.39999999999999997
+ }
+ }
+ },
"metadata": {
- "description": "Gemma 3n E2B IT is a multimodal, instruction-tuned model developed by Google DeepMind, designed to operate efficiently at an effective parameter size of 2B while leveraging a 6B architecture. Based on the MatFormer architecture, it supports nested submodels and modular composition via the Mix-and-Match framework. Gemma 3n models are optimized for low-resource deployment, offering 32K context length and strong multilingual and reasoning performance across common benchmarks. This variant is trained on a diverse corpus including code, math, web, and multimodal data.",
+ "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 70B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16217,35 +16573,43 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
"context_length": 8192,
- "max_completion_tokens": 2048,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-3n-e4b-it",
- "name": "Google: Gemma 3n 4B",
+ "id": "meta-llama/llama-3-8b-instruct",
+ "name": "Meta: Llama 3 8B Instruct",
"provider": "openrouter",
- "family": "google",
- "created_at": "2025-05-20 23:33:44 +0200",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "meta-llama",
+ "created_at": "2024-04-18 10:00:00 +1000",
+ "context_window": 8192,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -16257,18 +16621,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.02,
- "output_per_million": 0.04
+ "input_per_million": 0.03,
+ "output_per_million": 0.06
}
}
},
"metadata": {
- "description": "Gemma 3n E4B-it is optimized for efficient execution on mobile and low-resource devices, such as phones, laptops, and tablets. It supports multimodal inputs—including text, visual data, and audio—enabling diverse tasks such as text generation, speech recognition, translation, and image analysis. Leveraging innovations like Per-Layer Embedding (PLE) caching and the MatFormer architecture, Gemma 3n dynamically manages memory usage and computational load by selectively activating model parameters, significantly reducing runtime resource requirements.\n\nThis model supports a wide linguistic range (trained in over 140 languages) and features a flexible 32K token context window. Gemma 3n can selectively load parameters, optimizing memory and computational efficiency based on the task or device capabilities, making it well-suited for privacy-focused, offline-capable applications and on-device AI solutions. [Read more in the blog post](https://developers.googleblog.com/en/introducing-gemma-3n/)",
+ "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16277,12 +16643,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 8192,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
@@ -16293,21 +16659,25 @@
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
+ "seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "google/gemma-3n-e4b-it:free",
- "name": "Google: Gemma 3n 4B (free)",
+ "id": "meta-llama/llama-3.1-405b",
+ "name": "Meta: Llama 3.1 405B (base)",
"provider": "openrouter",
- "family": "google",
- "created_at": "2025-05-20 23:33:44 +0200",
- "context_window": 8192,
- "max_output_tokens": 2048,
+ "family": "meta-llama",
+ "created_at": "2024-08-02 10:00:00 +1000",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -16319,11 +16689,18 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 2.0
+ }
+ }
+ },
"metadata": {
- "description": "Gemma 3n E4B-it is optimized for efficient execution on mobile and low-resource devices, such as phones, laptops, and tablets. It supports multimodal inputs—including text, visual data, and audio—enabling diverse tasks such as text generation, speech recognition, translation, and image analysis. Leveraging innovations like Per-Layer Embedding (PLE) caching and the MatFormer architecture, Gemma 3n dynamically manages memory usage and computational load by selectively activating model parameters, significantly reducing runtime resource requirements.\n\nThis model supports a wide linguistic range (trained in over 140 languages) and features a flexible 32K token context window. Gemma 3n can selectively load parameters, optimizing memory and computational efficiency based on the task or device capabilities, making it well-suited for privacy-focused, offline-capable applications and on-device AI solutions. [Read more in the blog post](https://developers.googleblog.com/en/introducing-gemma-3n/)",
+ "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This is the base 405B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16332,35 +16709,40 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "none"
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 2048,
+ "context_length": 32768,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "repetition_penalty",
"seed",
"stop",
"temperature",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "gryphe/mythomax-l2-13b",
- "name": "MythoMax 13B",
+ "id": "meta-llama/llama-3.1-405b-instruct",
+ "name": "Meta: Llama 3.1 405B Instruct",
"provider": "openrouter",
- "family": "gryphe",
- "created_at": "2023-07-02 02:00:00 +0200",
- "context_window": 4096,
- "max_output_tokens": null,
+ "family": "meta-llama",
+ "created_at": "2024-07-23 10:00:00 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -16372,19 +16754,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.06,
- "output_per_million": 0.06
+ "input_per_million": 0.7999999999999999,
+ "output_per_million": 0.7999999999999999
}
}
},
"metadata": {
- "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge",
+ "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models including GPT-4o and Claude 3.5 Sonnet in evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16393,18 +16776,19 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama2",
- "instruct_type": "alpaca"
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 4096,
- "max_completion_tokens": null,
+ "context_length": 32768,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -16414,20 +16798,22 @@
"stop",
"structured_outputs",
"temperature",
- "top_a",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "inception/mercury",
- "name": "Inception: Mercury",
+ "id": "meta-llama/llama-3.1-405b-instruct:free",
+ "name": "Meta: Llama 3.1 405B Instruct (free)",
"provider": "openrouter",
- "family": "inception",
- "created_at": "2025-06-26 23:23:46 +0200",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "family": "meta-llama",
+ "created_at": "2024-07-23 10:00:00 +1000",
+ "context_window": 65536,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -16439,19 +16825,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.25,
- "output_per_million": 1.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Mercury is the first diffusion large language model (dLLM). Applying a breakthrough discrete diffusion approach, the model runs 5-10x faster than even speed optimized models like GPT-4.1 Nano and Claude 3.5 Haiku while matching their performance. Mercury's speed enables developers to provide responsive user experiences, including with voice agents, search interfaces, and chatbots. Read more in the blog post here. ",
+ "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models including GPT-4o and Claude 3.5 Sonnet in evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16460,12 +16838,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 16384,
+ "context_length": 65536,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -16477,20 +16855,18 @@
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "inception/mercury-coder",
- "name": "Inception: Mercury Coder",
+ "id": "meta-llama/llama-3.1-70b-instruct",
+ "name": "Meta: Llama 3.1 70B Instruct",
"provider": "openrouter",
- "family": "inception",
- "created_at": "2025-04-30 19:24:40 +0200",
- "context_window": 128000,
+ "family": "meta-llama",
+ "created_at": "2024-07-23 10:00:00 +1000",
+ "context_window": 131072,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
@@ -16504,18 +16880,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "output_per_million": 1.0
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.28
}
}
},
"metadata": {
- "description": "Mercury Coder is the first diffusion large language model (dLLM). Applying a breakthrough discrete diffusion approach, the model runs 5-10x faster than even speed optimized models like Claude 3.5 Haiku and GPT-4o Mini while matching their performance. Mercury Coder's speed means that developers can stay in the flow while coding, enjoying rapid chat-based iteration and responsive code completion suggestions. On Copilot Arena, Mercury Coder ranks 1st in speed and ties for 2nd in quality. Read more in the [blog post here](https://www.inceptionlabs.ai/introducing-mercury).",
+ "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 70B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16524,38 +16901,44 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 131072,
"max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
+ "seed",
"stop",
"structured_outputs",
"temperature",
"tool_choice",
"tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "infermatic/mn-inferor-12b",
- "name": "Infermatic: Mistral Nemo Inferor 12B",
+ "id": "meta-llama/llama-3.1-8b-instruct",
+ "name": "Meta: Llama 3.1 8B Instruct",
"provider": "openrouter",
- "family": "infermatic",
- "created_at": "2024-11-13 03:20:28 +0100",
- "context_window": 8192,
- "max_output_tokens": 8192,
+ "family": "meta-llama",
+ "created_at": "2024-07-23 10:00:00 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -16566,18 +16949,21 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.6,
- "output_per_million": 1.0
+ "input_per_million": 0.015,
+ "output_per_million": 0.02
}
}
},
"metadata": {
- "description": "Inferor 12B is a merge of top roleplay models, expert on immersive narratives and storytelling.\n\nThis model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [anthracite-org/magnum-v4-12b](https://openrouter.ai/anthracite-org/magnum-v4-72b) as a base.\n",
+ "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16586,147 +16972,170 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
- "instruct_type": "mistral"
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 8192,
+ "context_length": 131072,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "inflection/inflection-3-pi",
- "name": "Inflection: Inflection 3 Pi",
+ "id": "meta-llama/llama-3.2-11b-vision-instruct",
+ "name": "Meta: Llama 3.2 11B Vision Instruct",
"provider": "openrouter",
- "family": "inflection",
- "created_at": "2024-10-11 02:00:00 +0200",
- "context_window": 8000,
- "max_output_tokens": 1024,
+ "family": "meta-llama",
+ "created_at": "2024-09-25 10:00:00 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 0.049,
+ "output_per_million": 0.049
}
}
},
"metadata": {
- "description": "Inflection 3 Pi powers Inflection's [Pi](https://pi.ai) chatbot, including backstory, emotional intelligence, productivity, and safety. It has access to recent news, and excels in scenarios like customer support and roleplay.\n\nPi has been trained to mirror your tone and style, if you use more emojis, so will Pi! Try experimenting with various prompts and conversation styles.",
+ "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 8000,
- "max_completion_tokens": 1024,
+ "context_length": 131072,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "response_format",
+ "seed",
"stop",
+ "structured_outputs",
"temperature",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "inflection/inflection-3-productivity",
- "name": "Inflection: Inflection 3 Productivity",
+ "id": "meta-llama/llama-3.2-11b-vision-instruct:free",
+ "name": "Meta: Llama 3.2 11B Vision Instruct (free)",
"provider": "openrouter",
- "family": "inflection",
- "created_at": "2024-10-11 02:00:00 +0200",
- "context_window": 8000,
- "max_output_tokens": 1024,
+ "family": "meta-llama",
+ "created_at": "2024-09-25 10:00:00 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Inflection 3 Productivity is optimized for following instructions. It is better for tasks requiring JSON output or precise adherence to provided guidelines. It has access to recent news.\n\nFor emotional intelligence similar to Pi, see [Inflect 3 Pi](/inflection/inflection-3-pi)\n\nSee [Inflection's announcement](https://inflection.ai/blog/enterprise) for more details.",
+ "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 8000,
- "max_completion_tokens": 1024,
+ "context_length": 131072,
+ "max_completion_tokens": 2048,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"stop",
"temperature",
+ "top_k",
"top_p"
]
}
},
{
- "id": "liquid/lfm-3b",
- "name": "Liquid: LFM 3B",
+ "id": "meta-llama/llama-3.2-1b-instruct",
+ "name": "Meta: Llama 3.2 1B Instruct",
"provider": "openrouter",
- "family": "liquid",
- "created_at": "2025-01-25 13:01:41 +0100",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "meta-llama",
+ "created_at": "2024-09-25 10:00:00 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -16737,18 +17146,20 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.02,
- "output_per_million": 0.02
+ "input_per_million": 0.005,
+ "output_per_million": 0.01
}
}
},
"metadata": {
- "description": "Liquid's LFM 3B delivers incredible performance for its size. It positions itself as first place among 3B parameter transformers, hybrids, and RNN models It is also on par with Phi-3.5-mini on multiple benchmarks, while being 18.4% smaller.\n\nLFM-3B is the ideal choice for mobile and other edge text-based applications.\n\nSee the [launch announcement](https://www.liquid.ai/liquid-foundation-models) for benchmarks and more info.",
+ "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16757,37 +17168,41 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "chatml"
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "liquid/lfm-7b",
- "name": "Liquid: LFM 7B",
+ "id": "meta-llama/llama-3.2-3b-instruct",
+ "name": "Meta: Llama 3.2 3B Instruct",
"provider": "openrouter",
- "family": "liquid",
- "created_at": "2025-01-25 13:08:03 +0100",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "meta-llama",
+ "created_at": "2024-09-25 10:00:00 +1000",
+ "context_window": 20000,
+ "max_output_tokens": 20000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -16799,19 +17214,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01,
- "output_per_million": 0.01
+ "input_per_million": 0.003,
+ "output_per_million": 0.006
}
}
},
"metadata": {
- "description": "LFM-7B, a new best-in-class language model. LFM-7B is designed for exceptional chat capabilities, including languages like Arabic and Japanese. Powered by the Liquid Foundation Model (LFM) architecture, it exhibits unique features like low memory footprint and fast inference speed. \n\nLFM-7B is the world’s best-in-class multilingual language model in English, Arabic, and Japanese.\n\nSee the [launch announcement](https://www.liquid.ai/lfm-7b) for benchmarks and more info.",
+ "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16820,12 +17236,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "chatml"
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 20000,
+ "max_completion_tokens": 20000,
"is_moderated": false
},
"per_request_limits": null,
@@ -16840,7 +17256,10 @@
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -16848,13 +17267,13 @@
}
},
{
- "id": "mancer/weaver",
- "name": "Mancer: Weaver (alpha)",
+ "id": "meta-llama/llama-3.2-3b-instruct:free",
+ "name": "Meta: Llama 3.2 3B Instruct (free)",
"provider": "openrouter",
- "family": "mancer",
- "created_at": "2023-08-02 02:00:00 +0200",
- "context_window": 8000,
- "max_output_tokens": 1000,
+ "family": "meta-llama",
+ "created_at": "2024-09-25 10:00:00 +1000",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -16865,19 +17284,11 @@
]
},
"capabilities": [
- "streaming",
- "predicted_outputs"
+ "streaming"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 1.125,
- "output_per_million": 1.125
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "An attempt to recreate Claude-style verbosity, but don't expect the same level of coherence or memory. Meant for use in roleplay/narrative situations.",
+ "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -16886,43 +17297,39 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama2",
- "instruct_type": "alpaca"
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 8000,
- "max_completion_tokens": 1000,
+ "context_length": 131072,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
- "seed",
"stop",
"temperature",
- "top_a",
"top_k",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3-70b-instruct",
- "name": "Meta: Llama 3 70B Instruct",
+ "id": "meta-llama/llama-3.2-90b-vision-instruct",
+ "name": "Meta: Llama 3.2 90B Vision Instruct",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-04-18 02:00:00 +0200",
- "context_window": 8192,
- "max_output_tokens": 16384,
+ "created_at": "2024-09-25 10:00:00 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -16930,24 +17337,24 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "output_per_million": 0.39999999999999997
+ "input_per_million": 1.2,
+ "output_per_million": 1.2
}
}
},
"metadata": {
- "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 70B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "The Llama 90B Vision model is a top-tier, 90-billion-parameter multimodal model designed for the most challenging visual reasoning and language tasks. It offers unparalleled accuracy in image captioning, visual question answering, and advanced image-text comprehension. Pre-trained on vast multimodal datasets and fine-tuned with human feedback, the Llama 90B Vision is engineered to handle the most demanding image-based AI tasks.\n\nThis model is perfect for industries requiring cutting-edge multimodal AI capabilities, particularly those dealing with complex, real-time visual and textual analysis.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -16956,15 +17363,14 @@
"instruct_type": "llama3"
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 16384,
+ "context_length": 131072,
+ "max_completion_tokens": 2048,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -16973,21 +17379,18 @@
"seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3-8b-instruct",
- "name": "Meta: Llama 3 8B Instruct",
+ "id": "meta-llama/llama-3.3-70b-instruct",
+ "name": "Meta: Llama 3.3 70B Instruct",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-04-18 02:00:00 +0200",
- "context_window": 8192,
+ "created_at": "2024-12-07 04:28:57 +1100",
+ "context_window": 131072,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
@@ -17007,13 +17410,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.03,
- "output_per_million": 0.06
+ "input_per_million": 0.038000000000000006,
+ "output_per_million": 0.12
}
}
},
"metadata": {
- "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -17026,7 +17429,7 @@
"instruct_type": "llama3"
},
"top_provider": {
- "context_length": 8192,
+ "context_length": 131072,
"max_completion_tokens": 16384,
"is_moderated": false
},
@@ -17034,6 +17437,7 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -17041,21 +17445,23 @@
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.1-405b",
- "name": "Meta: Llama 3.1 405B (base)",
+ "id": "meta-llama/llama-3.3-70b-instruct:free",
+ "name": "Meta: Llama 3.3 70B Instruct (free)",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-08-02 02:00:00 +0200",
- "context_window": 32768,
+ "created_at": "2024-12-07 04:28:57 +1100",
+ "context_window": 65536,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -17068,18 +17474,12 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 2.0,
- "output_per_million": 2.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This is the base 405B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -17089,10 +17489,10 @@
"text"
],
"tokenizer": "Llama3",
- "instruct_type": "none"
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 65536,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -17100,28 +17500,27 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
- "seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.1-405b-instruct",
- "name": "Meta: Llama 3.1 405B Instruct",
+ "id": "meta-llama/llama-3.3-8b-instruct:free",
+ "name": "Meta: Llama 3.3 8B Instruct (free)",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-07-23 02:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": 16384,
+ "created_at": "2025-05-14 23:42:34 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 4028,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -17134,19 +17533,11 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.7999999999999999,
- "output_per_million": 0.7999999999999999
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models including GPT-4o and Claude 3.5 Sonnet in evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "A lightweight and ultra-fast variant of Llama 3.3 70B, for use when quick response times are needed most.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -17156,47 +17547,40 @@
"text"
],
"tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 16384,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 4028,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"repetition_penalty",
"response_format",
- "seed",
- "stop",
"structured_outputs",
"temperature",
"tool_choice",
"tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.1-405b-instruct:free",
- "name": "Meta: Llama 3.1 405B Instruct (free)",
+ "id": "meta-llama/llama-4-maverick",
+ "name": "Meta: Llama 4 Maverick",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-07-23 02:00:00 +0200",
- "context_window": 65536,
- "max_output_tokens": null,
+ "created_at": "2025-04-06 05:37:02 +1000",
+ "context_window": 1048576,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -17204,53 +17588,72 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
+ }
+ }
+ },
"metadata": {
- "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models including GPT-4o and Claude 3.5 Sonnet in evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Llama4",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 65536,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
+ "seed",
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.1-70b-instruct",
- "name": "Meta: Llama 3.1 70B Instruct",
+ "id": "meta-llama/llama-4-maverick:free",
+ "name": "Meta: Llama 4 Maverick (free)",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-07-23 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": 16384,
+ "created_at": "2025-04-06 05:37:02 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 4028,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -17259,69 +17662,55 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.28
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 70B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Llama4",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 16384,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 4028,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"repetition_penalty",
"response_format",
- "seed",
- "stop",
"structured_outputs",
"temperature",
"tool_choice",
"tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.1-8b-instruct",
- "name": "Meta: Llama 3.1 8B Instruct",
+ "id": "meta-llama/llama-4-scout",
+ "name": "Meta: Llama 4 Scout",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-07-23 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": 16384,
+ "created_at": "2025-04-06 05:31:59 +1000",
+ "context_window": 1048576,
+ "max_output_tokens": 1048576,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -17336,27 +17725,28 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.015,
- "output_per_million": 0.02
+ "input_per_million": 0.08,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Llama4",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 16384,
+ "context_length": 1048576,
+ "max_completion_tokens": 1048576,
"is_moderated": false
},
"per_request_limits": null,
@@ -17382,13 +17772,13 @@
}
},
{
- "id": "meta-llama/llama-3.2-11b-vision-instruct",
- "name": "Meta: Llama 3.2 11B Vision Instruct",
+ "id": "meta-llama/llama-4-scout:free",
+ "name": "Meta: Llama 4 Scout (free)",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-09-25 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": 16384,
+ "created_at": "2025-04-06 05:31:59 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 4028,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -17401,19 +17791,12 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.049,
- "output_per_million": 0.049
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
+ "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -17422,47 +17805,41 @@
],
"output_modalities": [
"text"
- ],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ ],
+ "tokenizer": "Llama4",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 16384,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 4028,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
"max_tokens",
- "min_p",
- "presence_penalty",
"repetition_penalty",
"response_format",
- "seed",
- "stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.2-11b-vision-instruct:free",
- "name": "Meta: Llama 3.2 11B Vision Instruct (free)",
+ "id": "meta-llama/llama-guard-2-8b",
+ "name": "Meta: LlamaGuard 2 8B",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-09-25 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": 2048,
+ "created_at": "2024-05-13 10:00:00 +1000",
+ "context_window": 8192,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -17472,24 +17849,30 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.19999999999999998
+ }
+ }
+ },
"metadata": {
- "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
+ "description": "This safeguard model has 8B parameters and is based on the Llama 3 family. Just like is predecessor, [LlamaGuard 1](https://huggingface.co/meta-llama/LlamaGuard-7b), it can do both prompt and response classification.\n\nLlamaGuard 2 acts as a normal LLM would, generating text that indicates whether the given input/output is safe/unsafe. If deemed unsafe, it will also share the content categories violated.\n\nFor best results, please use raw prompt input or the `/completions` endpoint, instead of the chat API.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
"tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "instruct_type": "none"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 2048,
+ "context_length": 8192,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -17508,13 +17891,13 @@
}
},
{
- "id": "meta-llama/llama-3.2-1b-instruct",
- "name": "Meta: Llama 3.2 1B Instruct",
+ "id": "meta-llama/llama-guard-3-8b",
+ "name": "Llama Guard 3 8B",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-09-25 02:00:00 +0200",
+ "created_at": "2025-02-13 10:01:58 +1100",
"context_window": 131072,
- "max_output_tokens": 16384,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -17532,13 +17915,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.005,
- "output_per_million": 0.01
+ "input_per_million": 0.02,
+ "output_per_million": 0.06
}
}
},
"metadata": {
- "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
+ "description": "Llama Guard 3 is a Llama-3.1-8B pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM – it generates text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 3 was aligned to safeguard against the MLCommons standardized hazards taxonomy and designed to support Llama 3.1 capabilities. Specifically, it provides content moderation in 8 languages, and was optimized to support safety and security for search and code interpreter tool calls.\n",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -17548,17 +17931,18 @@
"text"
],
"tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "instruct_type": "none"
},
"top_provider": {
"context_length": 131072,
- "max_completion_tokens": 16384,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -17566,7 +17950,6 @@
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"top_k",
"top_logprobs",
@@ -17575,16 +17958,17 @@
}
},
{
- "id": "meta-llama/llama-3.2-3b-instruct",
- "name": "Meta: Llama 3.2 3B Instruct",
+ "id": "meta-llama/llama-guard-4-12b",
+ "name": "Meta: Llama Guard 4 12B",
"provider": "openrouter",
"family": "meta-llama",
- "created_at": "2024-09-25 02:00:00 +0200",
- "context_window": 20000,
- "max_output_tokens": 20000,
+ "created_at": "2025-04-30 11:06:33 +1000",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -17593,41 +17977,40 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.003,
- "output_per_million": 0.006
+ "input_per_million": 0.18,
+ "output_per_million": 0.18
}
}
},
"metadata": {
- "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
+ "description": "Llama Guard 4 is a Llama 4 Scout-derived multimodal pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM—generating text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 4 was aligned to safeguard against the standardized MLCommons hazards taxonomy and designed to support multimodal Llama 4 capabilities. Specifically, it combines features from previous Llama Guard models, providing content moderation for English and multiple supported languages, along with enhanced capabilities to handle mixed text-and-image prompts, including multiple images. Additionally, Llama Guard 4 is integrated into the Llama Moderations API, extending robust safety classification to text and images.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
+ "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 20000,
- "max_completion_tokens": 20000,
+ "context_length": 163840,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -17635,23 +18018,19 @@
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.2-3b-instruct:free",
- "name": "Meta: Llama 3.2 3B Instruct (free)",
+ "id": "microsoft/mai-ds-r1",
+ "name": "Microsoft: MAI DS R1",
"provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2024-09-25 02:00:00 +0200",
- "context_window": 131072,
+ "family": "microsoft",
+ "created_at": "2025-04-21 10:08:20 +1000",
+ "context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -17663,11 +18042,19 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.1999188,
+ "output_per_million": 0.800064
+ }
+ }
+ },
"metadata": {
- "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
+ "description": "MAI-DS-R1 is a post-trained variant of DeepSeek-R1 developed by the Microsoft AI team to improve the model’s responsiveness on previously blocked topics while enhancing its safety profile. Built on top of DeepSeek-R1’s reasoning foundation, it integrates 110k examples from the Tulu-3 SFT dataset and 350k internally curated multilingual safety-alignment samples. The model retains strong reasoning, coding, and problem-solving capabilities, while unblocking a wide range of prompts previously restricted in R1.\n\nMAI-DS-R1 demonstrates improved performance on harm mitigation benchmarks and maintains competitive results across general reasoning tasks. It surpasses R1-1776 in satisfaction metrics for blocked queries and reduces leakage in harmful content categories. The model is based on a transformer MoE architecture and is suitable for general-purpose use cases, excluding high-stakes domains such as legal, medical, or autonomous systems.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -17676,39 +18063,46 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "DeepSeek",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 163840,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
+ "seed",
"stop",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.2-90b-vision-instruct",
- "name": "Meta: Llama 3.2 90B Vision Instruct",
+ "id": "microsoft/mai-ds-r1:free",
+ "name": "Microsoft: MAI DS R1 (free)",
"provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2024-09-25 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": 2048,
+ "family": "microsoft",
+ "created_at": "2025-04-21 10:08:20 +1000",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -17716,61 +18110,55 @@
},
"capabilities": [
"streaming",
- "structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 1.2,
- "output_per_million": 1.2
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "The Llama 90B Vision model is a top-tier, 90-billion-parameter multimodal model designed for the most challenging visual reasoning and language tasks. It offers unparalleled accuracy in image captioning, visual question answering, and advanced image-text comprehension. Pre-trained on vast multimodal datasets and fine-tuned with human feedback, the Llama 90B Vision is engineered to handle the most demanding image-based AI tasks.\n\nThis model is perfect for industries requiring cutting-edge multimodal AI capabilities, particularly those dealing with complex, real-time visual and textual analysis.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
+ "description": "MAI-DS-R1 is a post-trained variant of DeepSeek-R1 developed by the Microsoft AI team to improve the model’s responsiveness on previously blocked topics while enhancing its safety profile. Built on top of DeepSeek-R1’s reasoning foundation, it integrates 110k examples from the Tulu-3 SFT dataset and 350k internally curated multilingual safety-alignment samples. The model retains strong reasoning, coding, and problem-solving capabilities, while unblocking a wide range of prompts previously restricted in R1.\n\nMAI-DS-R1 demonstrates improved performance on harm mitigation benchmarks and maintains competitive results across general reasoning tasks. It surpasses R1-1776 in satisfaction metrics for blocked queries and reduces leakage in harmful content categories. The model is based on a transformer MoE architecture and is suitable for general-purpose use cases, excluding high-stakes domains such as legal, medical, or autonomous systems.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "DeepSeek",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 2048,
+ "context_length": 163840,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.3-70b-instruct",
- "name": "Meta: Llama 3.3 70B Instruct",
+ "id": "microsoft/phi-3-medium-128k-instruct",
+ "name": "Microsoft: Phi-3 Medium 128K Instruct",
"provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2024-12-06 18:28:57 +0100",
- "context_window": 131072,
- "max_output_tokens": 16384,
+ "family": "microsoft",
+ "created_at": "2024-05-24 10:00:00 +1000",
+ "context_window": 128000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -17782,20 +18170,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
- "predicted_outputs"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.038000000000000006,
- "output_per_million": 0.12
+ "input_per_million": 1.0,
+ "output_per_million": 1.0
}
}
},
"metadata": {
- "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)",
+ "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -17804,43 +18190,31 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Other",
+ "instruct_type": "phi3"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 16384,
+ "context_length": 128000,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "response_format",
- "seed",
- "stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-3.3-70b-instruct:free",
- "name": "Meta: Llama 3.3 70B Instruct (free)",
+ "id": "microsoft/phi-3-mini-128k-instruct",
+ "name": "Microsoft: Phi-3 Mini 128K Instruct",
"provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2024-12-06 18:28:57 +0100",
- "context_window": 65536,
+ "family": "microsoft",
+ "created_at": "2024-05-26 10:00:00 +1000",
+ "context_window": 128000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -17853,12 +18227,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "predicted_outputs"
+ "function_calling"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.09999999999999999
+ }
+ }
+ },
"metadata": {
- "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)",
+ "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -17867,44 +18247,36 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Other",
+ "instruct_type": "phi3"
},
"top_provider": {
- "context_length": 65536,
+ "context_length": 128000,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "stop",
"temperature",
"tool_choice",
"tools",
- "top_k",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-4-maverick",
- "name": "Meta: Llama 4 Maverick",
+ "id": "microsoft/phi-3.5-mini-128k-instruct",
+ "name": "Microsoft: Phi-3.5 Mini 128K Instruct",
"provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2025-04-05 21:37:02 +0200",
- "context_window": 1048576,
- "max_output_tokens": 16384,
+ "family": "microsoft",
+ "created_at": "2024-08-21 10:00:00 +1000",
+ "context_window": 128000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -17912,72 +18284,56 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
- "predicted_outputs"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.09999999999999999
}
}
},
"metadata": {
- "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.",
+ "description": "Phi-3.5 models are lightweight, state-of-the-art open models. These models were trained with Phi-3 datasets that include both synthetic data and the filtered, publicly available websites data, with a focus on high quality and reasoning-dense properties. Phi-3.5 Mini uses 3.8B parameters, and is a dense decoder-only transformer model using the same tokenizer as [Phi-3 Mini](/models/microsoft/phi-3-mini-128k-instruct).\n\nThe models underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures. When assessed against benchmarks that test common sense, language understanding, math, code, long context and logical reasoning, Phi-3.5 models showcased robust and state-of-the-art performance among models with less than 13 billion parameters.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama4",
- "instruct_type": null
+ "tokenizer": "Other",
+ "instruct_type": "phi3"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 16384,
+ "context_length": 128000,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "response_format",
- "seed",
- "stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-4-scout",
- "name": "Meta: Llama 4 Scout",
+ "id": "microsoft/phi-4",
+ "name": "Microsoft: Phi 4",
"provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2025-04-05 21:31:59 +0200",
- "context_window": 1048576,
- "max_output_tokens": 1048576,
+ "family": "microsoft",
+ "created_at": "2025-01-10 17:17:52 +1100",
+ "context_window": 16384,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -17985,35 +18341,33 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.08,
- "output_per_million": 0.3
+ "input_per_million": 0.06,
+ "output_per_million": 0.14
}
}
},
"metadata": {
- "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.",
+ "description": "[Microsoft Research](/microsoft) Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. \n\nAt 14 billion parameters, it was trained on a mix of high-quality synthetic datasets, data from curated websites, and academic materials. It has undergone careful improvement to follow instructions accurately and maintain strong safety standards. It works best with English language inputs.\n\nFor more information, please see [Phi-4 Technical Report](https://arxiv.org/pdf/2412.08905)\n",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama4",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 1048576,
+ "context_length": 16384,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -18030,8 +18384,6 @@
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -18039,17 +18391,18 @@
}
},
{
- "id": "meta-llama/llama-guard-2-8b",
- "name": "Meta: LlamaGuard 2 8B",
+ "id": "microsoft/phi-4-multimodal-instruct",
+ "name": "Microsoft: Phi 4 Multimodal Instruct",
"provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2024-05-13 02:00:00 +0200",
- "context_window": 8192,
+ "family": "microsoft",
+ "created_at": "2025-03-08 12:11:24 +1100",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -18057,42 +18410,44 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.19999999999999998
+ "input_per_million": 0.049999999999999996,
+ "output_per_million": 0.09999999999999999
}
}
},
"metadata": {
- "description": "This safeguard model has 8B parameters and is based on the Llama 3 family. Just like is predecessor, [LlamaGuard 1](https://huggingface.co/meta-llama/LlamaGuard-7b), it can do both prompt and response classification.\n\nLlamaGuard 2 acts as a normal LLM would, generating text that indicates whether the given input/output is safe/unsafe. If deemed unsafe, it will also share the content categories violated.\n\nFor best results, please use raw prompt input or the `/completions` endpoint, instead of the chat API.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "Phi-4 Multimodal Instruct is a versatile 5.6B parameter foundation model that combines advanced reasoning and instruction-following capabilities across both text and visual inputs, providing accurate text outputs. The unified architecture enables efficient, low-latency inference, suitable for edge and mobile deployments. Phi-4 Multimodal Instruct supports text inputs in multiple languages including Arabic, Chinese, English, French, German, Japanese, Spanish, and more, with visual input optimized primarily for English. It delivers impressive performance on multimodal tasks involving mathematical, scientific, and document reasoning, providing developers and enterprises a powerful yet compact model for sophisticated interactive applications. For more information, see the [Phi-4 Multimodal blog post](https://azure.microsoft.com/en-us/blog/empowering-innovation-the-next-generation-of-the-phi-family/).\n",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "none"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 8192,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
+ "seed",
"stop",
"temperature",
"top_k",
@@ -18101,12 +18456,12 @@
}
},
{
- "id": "meta-llama/llama-guard-3-8b",
- "name": "Llama Guard 3 8B",
+ "id": "microsoft/phi-4-reasoning-plus",
+ "name": "Microsoft: Phi 4 Reasoning Plus",
"provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2025-02-13 00:01:58 +0100",
- "context_window": 131072,
+ "family": "microsoft",
+ "created_at": "2025-05-02 06:22:41 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -18119,19 +18474,18 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.02,
- "output_per_million": 0.06
+ "input_per_million": 0.07,
+ "output_per_million": 0.35
}
}
},
"metadata": {
- "description": "Llama Guard 3 is a Llama-3.1-8B pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM – it generates text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 3 was aligned to safeguard against the MLCommons standardized hazards taxonomy and designed to support Llama 3.1 capabilities. Specifically, it provides content moderation in 8 languages, and was optimized to support safety and security for search and code interpreter tool calls.\n",
+ "description": "Phi-4-reasoning-plus is an enhanced 14B parameter model from Microsoft, fine-tuned from Phi-4 with additional reinforcement learning to boost accuracy on math, science, and code reasoning tasks. It uses the same dense decoder-only transformer architecture as Phi-4, but generates longer, more comprehensive outputs structured into a step-by-step reasoning trace and final answer.\n\nWhile it offers improved benchmark scores over Phi-4-reasoning across tasks like AIME, OmniMath, and HumanEvalPlus, its responses are typically ~50% longer, resulting in higher latency. Designed for English-only applications, it is well-suited for structured reasoning workflows where output quality takes priority over response speed.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18140,45 +18494,43 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "none"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
+ "include_reasoning",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
"response_format",
"seed",
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "meta-llama/llama-guard-4-12b",
- "name": "Meta: Llama Guard 4 12B",
+ "id": "microsoft/wizardlm-2-8x22b",
+ "name": "WizardLM-2 8x22B",
"provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2025-04-30 03:06:33 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "microsoft",
+ "created_at": "2024-04-16 10:00:00 +1000",
+ "context_window": 65536,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -18193,28 +18545,27 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.18,
- "output_per_million": 0.18
+ "input_per_million": 0.48,
+ "output_per_million": 0.48
}
}
},
"metadata": {
- "description": "Llama Guard 4 is a Llama 4 Scout-derived multimodal pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM—generating text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 4 was aligned to safeguard against the standardized MLCommons hazards taxonomy and designed to support multimodal Llama 4 capabilities. Specifically, it combines features from previous Llama Guard models, providing content moderation for English and multiple supported languages, along with enhanced capabilities to handle mixed text-and-image prompts, including multiple images. Additionally, Llama Guard 4 is integrated into the Llama Moderations API, extending robust safety classification to text and images.",
+ "description": "WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to leading proprietary models, and it consistently outperforms all existing state-of-the-art opensource models.\n\nIt is an instruct finetune of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Mistral",
+ "instruct_type": "vicuna"
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 65536,
+ "max_completion_tokens": 65536,
"is_moderated": false
},
"per_request_limits": null,
@@ -18235,80 +18586,69 @@
}
},
{
- "id": "microsoft/mai-ds-r1",
- "name": "Microsoft: MAI DS R1",
+ "id": "minimax/minimax-01",
+ "name": "MiniMax: MiniMax-01",
"provider": "openrouter",
- "family": "microsoft",
- "created_at": "2025-04-21 02:08:20 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "minimax",
+ "created_at": "2025-01-15 15:31:02 +1100",
+ "context_window": 1000192,
+ "max_output_tokens": 1000192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "predicted_outputs"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1999188,
- "output_per_million": 0.800064
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 1.1
}
}
},
"metadata": {
- "description": "MAI-DS-R1 is a post-trained variant of DeepSeek-R1 developed by the Microsoft AI team to improve the model’s responsiveness on previously blocked topics while enhancing its safety profile. Built on top of DeepSeek-R1’s reasoning foundation, it integrates 110k examples from the Tulu-3 SFT dataset and 350k internally curated multilingual safety-alignment samples. The model retains strong reasoning, coding, and problem-solving capabilities, while unblocking a wide range of prompts previously restricted in R1.\n\nMAI-DS-R1 demonstrates improved performance on harm mitigation benchmarks and maintains competitive results across general reasoning tasks. It surpasses R1-1776 in satisfaction metrics for blocked queries and reduces leakage in harmful content categories. The model is based on a transformer MoE architecture and is suitable for general-purpose use cases, excluding high-stakes domains such as legal, medical, or autonomous systems.",
+ "description": "MiniMax-01 is a combines MiniMax-Text-01 for text generation and MiniMax-VL-01 for image understanding. It has 456 billion parameters, with 45.9 billion parameters activated per inference, and can handle a context of up to 4 million tokens.\n\nThe text model adopts a hybrid architecture that combines Lightning Attention, Softmax Attention, and Mixture-of-Experts (MoE). The image model adopts the “ViT-MLP-LLM” framework and is trained on top of the text model.\n\nTo read more about the release, see: https://www.minimaxi.com/en/news/minimax-01-series-2",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 1000192,
+ "max_completion_tokens": 1000192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
- "seed",
- "stop",
"temperature",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "microsoft/mai-ds-r1:free",
- "name": "Microsoft: MAI DS R1 (free)",
+ "id": "minimax/minimax-m1",
+ "name": "MiniMax: MiniMax M1",
"provider": "openrouter",
- "family": "microsoft",
- "created_at": "2025-04-21 02:08:20 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "minimax",
+ "created_at": "2025-06-18 08:46:54 +1000",
+ "context_window": 1000000,
+ "max_output_tokens": 40000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -18320,11 +18660,19 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.3,
+ "output_per_million": 1.6500000000000001
+ }
+ }
+ },
"metadata": {
- "description": "MAI-DS-R1 is a post-trained variant of DeepSeek-R1 developed by the Microsoft AI team to improve the model’s responsiveness on previously blocked topics while enhancing its safety profile. Built on top of DeepSeek-R1’s reasoning foundation, it integrates 110k examples from the Tulu-3 SFT dataset and 350k internally curated multilingual safety-alignment samples. The model retains strong reasoning, coding, and problem-solving capabilities, while unblocking a wide range of prompts previously restricted in R1.\n\nMAI-DS-R1 demonstrates improved performance on harm mitigation benchmarks and maintains competitive results across general reasoning tasks. It surpasses R1-1776 in satisfaction metrics for blocked queries and reduces leakage in harmful content categories. The model is based on a transformer MoE architecture and is suitable for general-purpose use cases, excluding high-stakes domains such as legal, medical, or autonomous systems.",
+ "description": "MiniMax-M1 is a large-scale, open-weight reasoning model designed for extended context and high-efficiency inference. It leverages a hybrid Mixture-of-Experts (MoE) architecture paired with a custom \"lightning attention\" mechanism, allowing it to process long sequences—up to 1 million tokens—while maintaining competitive FLOP efficiency. With 456 billion total parameters and 45.9B active per token, this variant is optimized for complex, multi-step reasoning tasks.\n\nTrained via a custom reinforcement learning pipeline (CISPO), M1 excels in long-context understanding, software engineering, agentic tool use, and mathematical reasoning. Benchmarks show strong performance across FullStackBench, SWE-bench, MATH, GPQA, and TAU-Bench, often outperforming other open models like DeepSeek R1 and Qwen3-235B.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18333,12 +18681,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 1000000,
+ "max_completion_tokens": 40000,
"is_moderated": false
},
"per_request_limits": null,
@@ -18346,7 +18694,6 @@
"frequency_penalty",
"include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -18354,20 +18701,22 @@
"repetition_penalty",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "microsoft/phi-3-medium-128k-instruct",
- "name": "Microsoft: Phi-3 Medium 128K Instruct",
+ "id": "mistralai/codestral-2501",
+ "name": "Mistral: Codestral 2501",
"provider": "openrouter",
- "family": "microsoft",
- "created_at": "2024-05-24 02:00:00 +0200",
- "context_window": 128000,
+ "family": "mistralai",
+ "created_at": "2025-01-15 09:58:42 +1100",
+ "context_window": 262144,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -18380,18 +18729,19 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.0,
- "output_per_million": 1.0
+ "input_per_million": 0.3,
+ "output_per_million": 0.8999999999999999
}
}
},
"metadata": {
- "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).",
+ "description": "[Mistral](/mistralai)'s cutting-edge language model for coding. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. \n\nLearn more on their blog post: https://mistral.ai/news/codestral-2501/",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18400,17 +18750,23 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "phi3"
+ "tokenizer": "Mistral",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 262144,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"max_tokens",
+ "presence_penalty",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -18419,12 +18775,12 @@
}
},
{
- "id": "microsoft/phi-3-mini-128k-instruct",
- "name": "Microsoft: Phi-3 Mini 128K Instruct",
+ "id": "mistralai/codestral-2508",
+ "name": "Mistral: Codestral 2508",
"provider": "openrouter",
- "family": "microsoft",
- "created_at": "2024-05-26 02:00:00 +0200",
- "context_window": 128000,
+ "family": "mistralai",
+ "created_at": "2025-08-02 06:20:30 +1000",
+ "context_window": 256000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -18437,18 +18793,19 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.09999999999999999
+ "input_per_million": 0.3,
+ "output_per_million": 0.8999999999999999
}
}
},
"metadata": {
- "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.",
+ "description": "Mistral's cutting-edge language model for coding released end of July 2025. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation.\n\n[Blog Post](https://mistral.ai/news/codestral-25-08)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18457,17 +18814,23 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "phi3"
+ "tokenizer": "Mistral",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 256000,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"max_tokens",
+ "presence_penalty",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -18476,12 +18839,12 @@
}
},
{
- "id": "microsoft/phi-3.5-mini-128k-instruct",
- "name": "Microsoft: Phi-3.5 Mini 128K Instruct",
+ "id": "mistralai/devstral-medium",
+ "name": "Mistral: Devstral Medium",
"provider": "openrouter",
- "family": "microsoft",
- "created_at": "2024-08-21 02:00:00 +0200",
- "context_window": 128000,
+ "family": "mistralai",
+ "created_at": "2025-07-11 01:28:41 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -18494,18 +18857,19 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.09999999999999999
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 2.0
}
}
},
"metadata": {
- "description": "Phi-3.5 models are lightweight, state-of-the-art open models. These models were trained with Phi-3 datasets that include both synthetic data and the filtered, publicly available websites data, with a focus on high quality and reasoning-dense properties. Phi-3.5 Mini uses 3.8B parameters, and is a dense decoder-only transformer model using the same tokenizer as [Phi-3 Mini](/models/microsoft/phi-3-mini-128k-instruct).\n\nThe models underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures. When assessed against benchmarks that test common sense, language understanding, math, code, long context and logical reasoning, Phi-3.5 models showcased robust and state-of-the-art performance among models with less than 13 billion parameters.",
+ "description": "Devstral Medium is a high-performance code generation and agentic reasoning model developed jointly by Mistral AI and All Hands AI. Positioned as a step up from Devstral Small, it achieves 61.6% on SWE-Bench Verified, placing it ahead of Gemini 2.5 Pro and GPT-4.1 in code-related tasks, at a fraction of the cost. It is designed for generalization across prompt styles and tool use in code agents and frameworks.\n\nDevstral Medium is available via API only (not open-weight), and supports enterprise deployment on private infrastructure, with optional fine-tuning capabilities.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18514,17 +18878,23 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "phi3"
+ "tokenizer": "Mistral",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"max_tokens",
+ "presence_penalty",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -18533,12 +18903,12 @@
}
},
{
- "id": "microsoft/phi-4",
- "name": "Microsoft: Phi 4",
+ "id": "mistralai/devstral-small",
+ "name": "Mistral: Devstral Small 1.1",
"provider": "openrouter",
- "family": "microsoft",
- "created_at": "2025-01-10 07:17:52 +0100",
- "context_window": 16384,
+ "family": "mistralai",
+ "created_at": "2025-07-11 01:19:11 +1000",
+ "context_window": 128000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -18551,19 +18921,19 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.06,
- "output_per_million": 0.14
+ "input_per_million": 0.07,
+ "output_per_million": 0.28
}
}
},
"metadata": {
- "description": "[Microsoft Research](/microsoft) Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. \n\nAt 14 billion parameters, it was trained on a mix of high-quality synthetic datasets, data from curated websites, and academic materials. It has undergone careful improvement to follow instructions accurately and maintain strong safety standards. It works best with English language inputs.\n\nFor more information, please see [Phi-4 Technical Report](https://arxiv.org/pdf/2412.08905)\n",
+ "description": "Devstral Small 1.1 is a 24B parameter open-weight language model for software engineering agents, developed by Mistral AI in collaboration with All Hands AI. Finetuned from Mistral Small 3.1 and released under the Apache 2.0 license, it features a 128k token context window and supports both Mistral-style function calling and XML output formats.\n\nDesigned for agentic coding workflows, Devstral Small 1.1 is optimized for tasks such as codebase exploration, multi-file edits, and integration into autonomous development agents like OpenHands and Cline. It achieves 53.6% on SWE-Bench Verified, surpassing all other open models on this benchmark, while remaining lightweight enough to run on a single 4090 GPU or Apple silicon machine. The model uses a Tekken tokenizer with a 131k vocabulary and is deployable via vLLM, Transformers, Ollama, LM Studio, and other OpenAI-compatible runtimes.\n",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18572,19 +18942,17 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Mistral",
"instruct_type": null
},
"top_provider": {
- "context_length": 16384,
+ "context_length": 128000,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -18594,25 +18962,25 @@
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "microsoft/phi-4-multimodal-instruct",
- "name": "Microsoft: Phi 4 Multimodal Instruct",
+ "id": "mistralai/devstral-small-2505",
+ "name": "Mistral: Devstral Small 2505",
"provider": "openrouter",
- "family": "microsoft",
- "created_at": "2025-03-08 02:11:24 +0100",
+ "family": "mistralai",
+ "created_at": "2025-05-22 00:22:59 +1000",
"context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -18620,28 +18988,29 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.049999999999999996,
- "output_per_million": 0.09999999999999999
+ "input_per_million": 0.01999188,
+ "output_per_million": 0.0800064
}
}
},
"metadata": {
- "description": "Phi-4 Multimodal Instruct is a versatile 5.6B parameter foundation model that combines advanced reasoning and instruction-following capabilities across both text and visual inputs, providing accurate text outputs. The unified architecture enables efficient, low-latency inference, suitable for edge and mobile deployments. Phi-4 Multimodal Instruct supports text inputs in multiple languages including Arabic, Chinese, English, French, German, Japanese, Spanish, and more, with visual input optimized primarily for English. It delivers impressive performance on multimodal tasks involving mathematical, scientific, and document reasoning, providing developers and enterprises a powerful yet compact model for sophisticated interactive applications. For more information, see the [Phi-4 Multimodal blog post](https://azure.microsoft.com/en-us/blog/empowering-innovation-the-next-generation-of-the-phi-family/).\n",
+ "description": "Devstral-Small-2505 is a 24B parameter agentic LLM fine-tuned from Mistral-Small-3.1, jointly developed by Mistral AI and All Hands AI for advanced software engineering tasks. It is optimized for codebase exploration, multi-file editing, and integration into coding agents, achieving state-of-the-art results on SWE-Bench Verified (46.8%).\n\nDevstral supports a 128k context window and uses a custom Tekken tokenizer. It is text-only, with the vision encoder removed, and is suitable for local deployment on high-end consumer hardware (e.g., RTX 4090, 32GB RAM Macs). Devstral is best used in agentic workflows via the OpenHands scaffold and is compatible with inference frameworks like vLLM, Transformers, and Ollama. It is released under the Apache 2.0 license.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Mistral",
"instruct_type": null
},
"top_provider": {
@@ -18652,6 +19021,8 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -18659,18 +19030,22 @@
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "microsoft/phi-4-reasoning-plus",
- "name": "Microsoft: Phi 4 Reasoning Plus",
+ "id": "mistralai/devstral-small-2505:free",
+ "name": "Mistral: Devstral Small 2505 (free)",
"provider": "openrouter",
- "family": "microsoft",
- "created_at": "2025-05-01 22:22:41 +0200",
+ "family": "mistralai",
+ "created_at": "2025-05-22 00:22:59 +1000",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -18684,18 +19059,12 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.07,
- "output_per_million": 0.35
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Phi-4-reasoning-plus is an enhanced 14B parameter model from Microsoft, fine-tuned from Phi-4 with additional reinforcement learning to boost accuracy on math, science, and code reasoning tasks. It uses the same dense decoder-only transformer architecture as Phi-4, but generates longer, more comprehensive outputs structured into a step-by-step reasoning trace and final answer.\n\nWhile it offers improved benchmark scores over Phi-4-reasoning across tasks like AIME, OmniMath, and HumanEvalPlus, its responses are typically ~50% longer, resulting in higher latency. Designed for English-only applications, it is well-suited for structured reasoning workflows where output quality takes priority over response speed.",
+ "description": "Devstral-Small-2505 is a 24B parameter agentic LLM fine-tuned from Mistral-Small-3.1, jointly developed by Mistral AI and All Hands AI for advanced software engineering tasks. It is optimized for codebase exploration, multi-file editing, and integration into coding agents, achieving state-of-the-art results on SWE-Bench Verified (46.8%).\n\nDevstral supports a 128k context window and uses a custom Tekken tokenizer. It is text-only, with the vision encoder removed, and is suitable for local deployment on high-end consumer hardware (e.g., RTX 4090, 32GB RAM Macs). Devstral is best used in agentic workflows via the OpenHands scaffold and is compatible with inference frameworks like vLLM, Transformers, and Ollama. It is released under the Apache 2.0 license.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18704,7 +19073,7 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Mistral",
"instruct_type": null
},
"top_provider": {
@@ -18715,29 +19084,31 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "microsoft/wizardlm-2-8x22b",
- "name": "WizardLM-2 8x22B",
+ "id": "mistralai/magistral-medium-2506",
+ "name": "Mistral: Magistral Medium 2506",
"provider": "openrouter",
- "family": "microsoft",
- "created_at": "2024-04-16 02:00:00 +0200",
- "context_window": 65536,
- "max_output_tokens": 65536,
+ "family": "mistralai",
+ "created_at": "2025-06-08 13:40:54 +1000",
+ "context_window": 40960,
+ "max_output_tokens": 40000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -18749,19 +19120,19 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.48,
- "output_per_million": 0.48
+ "input_per_million": 2.0,
+ "output_per_million": 5.0
}
}
},
"metadata": {
- "description": "WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to leading proprietary models, and it consistently outperforms all existing state-of-the-art opensource models.\n\nIt is an instruct finetune of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe",
+ "description": "Magistral is Mistral's first reasoning model. It is ideal for general purpose use requiring longer thought processing and better accuracy than with non-reasoning LLMs. From legal research and financial forecasting to software development and creative storytelling — this model solves multi-step challenges where transparency and precision are critical.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18771,93 +19142,104 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": "vicuna"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 65536,
- "max_completion_tokens": 65536,
+ "context_length": 40960,
+ "max_completion_tokens": 40000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
+ "include_reasoning",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "reasoning",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "minimax/minimax-01",
- "name": "MiniMax: MiniMax-01",
+ "id": "mistralai/magistral-medium-2506:thinking",
+ "name": "Mistral: Magistral Medium 2506 (thinking)",
"provider": "openrouter",
- "family": "minimax",
- "created_at": "2025-01-15 05:31:02 +0100",
- "context_window": 1000192,
- "max_output_tokens": 1000192,
+ "family": "mistralai",
+ "created_at": "2025-06-08 13:40:54 +1000",
+ "context_window": 40960,
+ "max_output_tokens": 40000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 1.1
+ "input_per_million": 2.0,
+ "output_per_million": 5.0
}
}
},
"metadata": {
- "description": "MiniMax-01 is a combines MiniMax-Text-01 for text generation and MiniMax-VL-01 for image understanding. It has 456 billion parameters, with 45.9 billion parameters activated per inference, and can handle a context of up to 4 million tokens.\n\nThe text model adopts a hybrid architecture that combines Lightning Attention, Softmax Attention, and Mixture-of-Experts (MoE). The image model adopts the “ViT-MLP-LLM” framework and is trained on top of the text model.\n\nTo read more about the release, see: https://www.minimaxi.com/en/news/minimax-01-series-2",
+ "description": "Magistral is Mistral's first reasoning model. It is ideal for general purpose use requiring longer thought processing and better accuracy than with non-reasoning LLMs. From legal research and financial forecasting to software development and creative storytelling — this model solves multi-step challenges where transparency and precision are critical.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Mistral",
"instruct_type": null
},
"top_provider": {
- "context_length": 1000192,
- "max_completion_tokens": 1000192,
+ "context_length": 40960,
+ "max_completion_tokens": 40000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
"max_tokens",
+ "presence_penalty",
+ "reasoning",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "minimax/minimax-m1",
- "name": "MiniMax: MiniMax M1",
+ "id": "mistralai/magistral-small-2506",
+ "name": "Mistral: Magistral Small 2506",
"provider": "openrouter",
- "family": "minimax",
- "created_at": "2025-06-18 00:46:54 +0200",
- "context_window": 1000000,
+ "family": "mistralai",
+ "created_at": "2025-06-11 01:32:41 +1000",
+ "context_window": 40000,
"max_output_tokens": 40000,
"knowledge_cutoff": null,
"modalities": {
@@ -18871,18 +19253,18 @@
"capabilities": [
"streaming",
"function_calling",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "output_per_million": 1.6500000000000001
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
}
}
},
"metadata": {
- "description": "MiniMax-M1 is a large-scale, open-weight reasoning model designed for extended context and high-efficiency inference. It leverages a hybrid Mixture-of-Experts (MoE) architecture paired with a custom \"lightning attention\" mechanism, allowing it to process long sequences—up to 1 million tokens—while maintaining competitive FLOP efficiency. With 456 billion total parameters and 45.9B active per token, this variant is optimized for complex, multi-step reasoning tasks.\n\nTrained via a custom reinforcement learning pipeline (CISPO), M1 excels in long-context understanding, software engineering, agentic tool use, and mathematical reasoning. Benchmarks show strong performance across FullStackBench, SWE-bench, MATH, GPQA, and TAU-Bench, often outperforming other open models like DeepSeek R1 and Qwen3-235B.",
+ "description": "Magistral Small is a 24B parameter instruction-tuned model based on Mistral-Small-3.1 (2503), enhanced through supervised fine-tuning on traces from Magistral Medium and further refined via reinforcement learning. It is optimized for reasoning and supports a wide multilingual range, including over 20 languages.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18891,11 +19273,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Mistral",
"instruct_type": null
},
"top_provider": {
- "context_length": 1000000,
+ "context_length": 40000,
"max_completion_tokens": 40000,
"is_moderated": false
},
@@ -18903,30 +19285,27 @@
"supported_parameters": [
"frequency_penalty",
"include_reasoning",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
"reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
"tool_choice",
"tools",
- "top_k",
"top_p"
]
}
},
{
- "id": "mistralai/codestral-2501",
- "name": "Mistral: Codestral 2501",
+ "id": "mistralai/ministral-3b",
+ "name": "Mistral: Ministral 3B",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-01-14 23:58:42 +0100",
- "context_window": 262144,
+ "created_at": "2024-10-17 11:00:00 +1100",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -18939,19 +19318,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "output_per_million": 0.8999999999999999
+ "input_per_million": 0.04,
+ "output_per_million": 0.04
}
}
},
"metadata": {
- "description": "[Mistral](/mistralai)'s cutting-edge language model for coding. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. \n\nLearn more on their blog post: https://mistral.ai/news/codestral-2501/",
+ "description": "Ministral 3B is a 3B parameter model optimized for on-device and edge computing. It excels in knowledge, commonsense reasoning, and function-calling, outperforming larger models like Mistral 7B on most benchmarks. Supporting up to 128k context length, it’s ideal for orchestrating agentic workflows and specialist tasks with efficient inference.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -18964,7 +19342,7 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 262144,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -18978,19 +19356,17 @@
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_p"
]
}
},
{
- "id": "mistralai/codestral-2508",
- "name": "Mistral: Codestral 2508",
+ "id": "mistralai/ministral-8b",
+ "name": "Mistral: Ministral 8B",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-08-01 22:20:30 +0200",
- "context_window": 256000,
+ "created_at": "2024-10-17 11:00:00 +1100",
+ "context_window": 128000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -19009,13 +19385,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "output_per_million": 0.8999999999999999
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.09999999999999999
}
}
},
"metadata": {
- "description": "Mistral's cutting-edge language model for coding released end of July 2025. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation.\n\n[Blog Post](https://mistral.ai/news/codestral-25-08)",
+ "description": "Ministral 8B is an 8B parameter model featuring a unique interleaved sliding-window attention pattern for faster, memory-efficient inference. Designed for edge use cases, it supports up to 128k context length and excels in knowledge and reasoning tasks. It outperforms peers in the sub-10B category, making it perfect for low-latency, privacy-first applications.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19028,7 +19404,7 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 256000,
+ "context_length": 128000,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -19049,13 +19425,13 @@
}
},
{
- "id": "mistralai/devstral-medium",
- "name": "Mistral: Devstral Medium",
+ "id": "mistralai/mistral-7b-instruct",
+ "name": "Mistral: Mistral 7B Instruct",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-07-10 17:28:41 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "created_at": "2024-05-27 10:00:00 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -19068,18 +19444,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 2.0
+ "input_per_million": 0.028,
+ "output_per_million": 0.054
}
}
},
"metadata": {
- "description": "Devstral Medium is a high-performance code generation and agentic reasoning model developed jointly by Mistral AI and All Hands AI. Positioned as a step up from Devstral Small, it achieves 61.6% on SWE-Bench Verified, placing it ahead of Gemini 2.5 Pro and GPT-4.1 in code-related tasks, at a fraction of the cost. It is designed for generalization across prompt styles and tool use in code agents and frameworks.\n\nDevstral Medium is available via API only (not open-weight), and supports enterprise deployment on private infrastructure, with optional fine-tuning capabilities.",
+ "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19089,36 +19466,40 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": null
+ "instruct_type": "mistral"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
+ "context_length": 32768,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
+ "top_k",
"top_p"
]
}
},
{
- "id": "mistralai/devstral-small",
- "name": "Mistral: Devstral Small 1.1",
+ "id": "mistralai/mistral-7b-instruct-v0.1",
+ "name": "Mistral: Mistral 7B Instruct v0.1",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-07-10 17:19:11 +0200",
- "context_window": 128000,
+ "created_at": "2023-09-28 10:00:00 +1000",
+ "context_window": 2824,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -19132,18 +19513,18 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.07,
- "output_per_million": 0.28
+ "input_per_million": 0.11,
+ "output_per_million": 0.19
}
}
},
"metadata": {
- "description": "Devstral Small 1.1 is a 24B parameter open-weight language model for software engineering agents, developed by Mistral AI in collaboration with All Hands AI. Finetuned from Mistral Small 3.1 and released under the Apache 2.0 license, it features a 128k token context window and supports both Mistral-style function calling and XML output formats.\n\nDesigned for agentic coding workflows, Devstral Small 1.1 is optimized for tasks such as codebase exploration, multi-file edits, and integration into autonomous development agents like OpenHands and Cline. It achieves 53.6% on SWE-Bench Verified, surpassing all other open models on this benchmark, while remaining lightweight enough to run on a single 4090 GPU or Apple silicon machine. The model uses a Tekken tokenizer with a 131k vocabulary and is deployable via vLLM, Transformers, Ollama, LM Studio, and other OpenAI-compatible runtimes.\n",
+ "description": "A 7.3B parameter model that outperforms Llama 2 13B on all benchmarks, with optimizations for speed and context length.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19153,24 +19534,23 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": null
+ "instruct_type": "mistral"
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 2824,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -19180,13 +19560,13 @@
}
},
{
- "id": "mistralai/devstral-small-2505",
- "name": "Mistral: Devstral Small 2505",
+ "id": "mistralai/mistral-7b-instruct-v0.3",
+ "name": "Mistral: Mistral 7B Instruct v0.3",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-05-21 16:22:59 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "created_at": "2024-05-27 10:00:00 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -19205,13 +19585,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01999188,
- "output_per_million": 0.0800064
+ "input_per_million": 0.028,
+ "output_per_million": 0.054
}
}
},
"metadata": {
- "description": "Devstral-Small-2505 is a 24B parameter agentic LLM fine-tuned from Mistral-Small-3.1, jointly developed by Mistral AI and All Hands AI for advanced software engineering tasks. It is optimized for codebase exploration, multi-file editing, and integration into coding agents, achieving state-of-the-art results on SWE-Bench Verified (46.8%).\n\nDevstral supports a 128k context window and uses a custom Tekken tokenizer. It is text-only, with the vision encoder removed, and is suitable for local deployment on high-end consumer hardware (e.g., RTX 4090, 32GB RAM Macs). Devstral is best used in agentic workflows via the OpenHands scaffold and is compatible with inference frameworks like vLLM, Transformers, and Ollama. It is released under the Apache 2.0 license.",
+ "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct v0.2](/models/mistralai/mistral-7b-instruct-v0.2), with the following changes:\n\n- Extended vocabulary to 32768\n- Supports v3 Tokenizer\n- Supports function calling\n\nNOTE: Support for function calling depends on the provider.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19221,11 +19601,11 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": null
+ "instruct_type": "mistral"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
+ "context_length": 32768,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
@@ -19240,24 +19620,22 @@
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/devstral-small-2505:free",
- "name": "Mistral: Devstral Small 2505 (free)",
+ "id": "mistralai/mistral-7b-instruct:free",
+ "name": "Mistral: Mistral 7B Instruct (free)",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-05-21 16:22:59 +0200",
+ "created_at": "2024-05-27 10:00:00 +1000",
"context_window": 32768,
- "max_output_tokens": null,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -19270,11 +19648,11 @@
"capabilities": [
"streaming",
"function_calling",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {},
"metadata": {
- "description": "Devstral-Small-2505 is a 24B parameter agentic LLM fine-tuned from Mistral-Small-3.1, jointly developed by Mistral AI and All Hands AI for advanced software engineering tasks. It is optimized for codebase exploration, multi-file editing, and integration into coding agents, achieving state-of-the-art results on SWE-Bench Verified (46.8%).\n\nDevstral supports a 128k context window and uses a custom Tekken tokenizer. It is text-only, with the vision encoder removed, and is suitable for local deployment on high-end consumer hardware (e.g., RTX 4090, 32GB RAM Macs). Devstral is best used in agentic workflows via the OpenHands scaffold and is compatible with inference frameworks like vLLM, Transformers, and Ollama. It is released under the Apache 2.0 license.",
+ "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19284,41 +19662,39 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": null
+ "instruct_type": "mistral"
},
"top_provider": {
"context_length": 32768,
- "max_completion_tokens": null,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
"temperature",
"tool_choice",
"tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/magistral-medium-2506",
- "name": "Mistral: Magistral Medium 2506",
+ "id": "mistralai/mistral-large",
+ "name": "Mistral Large",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-06-08 05:40:54 +0200",
- "context_window": 40960,
- "max_output_tokens": 40000,
+ "created_at": "2024-02-26 11:00:00 +1100",
+ "context_window": 128000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -19337,12 +19713,12 @@
"text_tokens": {
"standard": {
"input_per_million": 2.0,
- "output_per_million": 5.0
+ "output_per_million": 6.0
}
}
},
"metadata": {
- "description": "Magistral is Mistral's first reasoning model. It is ideal for general purpose use requiring longer thought processing and better accuracy than with non-reasoning LLMs. From legal research and financial forecasting to software development and creative storytelling — this model solves multi-step challenges where transparency and precision are critical.",
+ "description": "This is Mistral AI's flagship model, Mistral Large 2 (version `mistral-large-2407`). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19355,17 +19731,15 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 40960,
- "max_completion_tokens": 40000,
+ "context_length": 128000,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"max_tokens",
"presence_penalty",
- "reasoning",
"response_format",
"seed",
"stop",
@@ -19378,13 +19752,13 @@
}
},
{
- "id": "mistralai/magistral-medium-2506:thinking",
- "name": "Mistral: Magistral Medium 2506 (thinking)",
+ "id": "mistralai/mistral-large-2407",
+ "name": "Mistral Large 2407",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-06-08 05:40:54 +0200",
- "context_window": 40960,
- "max_output_tokens": 40000,
+ "created_at": "2024-11-19 12:06:55 +1100",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -19403,12 +19777,12 @@
"text_tokens": {
"standard": {
"input_per_million": 2.0,
- "output_per_million": 5.0
+ "output_per_million": 6.0
}
}
},
"metadata": {
- "description": "Magistral is Mistral's first reasoning model. It is ideal for general purpose use requiring longer thought processing and better accuracy than with non-reasoning LLMs. From legal research and financial forecasting to software development and creative storytelling — this model solves multi-step challenges where transparency and precision are critical.",
+ "description": "This is Mistral AI's flagship model, Mistral Large 2 (version mistral-large-2407). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.\n",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19421,17 +19795,15 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 40960,
- "max_completion_tokens": 40000,
+ "context_length": 131072,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"max_tokens",
"presence_penalty",
- "reasoning",
"response_format",
"seed",
"stop",
@@ -19444,13 +19816,13 @@
}
},
{
- "id": "mistralai/magistral-small-2506",
- "name": "Mistral: Magistral Small 2506",
+ "id": "mistralai/mistral-large-2411",
+ "name": "Mistral Large 2411",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-06-10 17:32:41 +0200",
- "context_window": 40000,
- "max_output_tokens": 40000,
+ "created_at": "2024-11-19 12:11:25 +1100",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -19468,13 +19840,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.0,
+ "output_per_million": 6.0
}
}
},
"metadata": {
- "description": "Magistral Small is a 24B parameter instruction-tuned model based on Mistral-Small-3.1 (2503), enhanced through supervised fine-tuning on traces from Magistral Medium and further refined via reinforcement learning. It is optimized for reasoning and supports a wide multilingual range, including over 20 languages.",
+ "description": "Mistral Large 2 2411 is an update of [Mistral Large 2](/mistralai/mistral-large) released together with [Pixtral Large 2411](/mistralai/pixtral-large-2411)\n\nIt provides a significant upgrade on the previous [Mistral Large 24.07](/mistralai/mistral-large-2407), with notable improvements in long context understanding, a new system prompt, and more accurate function calling.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19487,17 +19859,81 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 40000,
- "max_completion_tokens": 40000,
+ "context_length": 131072,
+ "max_completion_tokens": null,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "frequency_penalty",
+ "max_tokens",
+ "presence_penalty",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
+ "temperature",
+ "tool_choice",
+ "tools",
+ "top_p"
+ ]
+ }
+ },
+ {
+ "id": "mistralai/mistral-medium-3",
+ "name": "Mistral: Mistral Medium 3",
+ "provider": "openrouter",
+ "family": "mistralai",
+ "created_at": "2025-05-08 00:15:41 +1000",
+ "context_window": 131072,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 2.0
+ }
+ }
+ },
+ "metadata": {
+ "description": "Mistral Medium 3 is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances state-of-the-art reasoning and multimodal performance with 8× lower cost compared to traditional large models, making it suitable for scalable deployments across professional and industrial use cases.\n\nThe model excels in domains such as coding, STEM reasoning, and enterprise adaptation. It supports hybrid, on-prem, and in-VPC deployments and is optimized for integration into custom workflows. Mistral Medium 3 offers competitive accuracy relative to larger models like Claude Sonnet 3.5/3.7, Llama 4 Maverick, and Command R+, while maintaining broad compatibility across cloud environments.",
+ "architecture": {
+ "modality": "text+image->text",
+ "input_modalities": [
+ "text",
+ "image"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Mistral",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 131072,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"max_tokens",
"presence_penalty",
- "reasoning",
"response_format",
"seed",
"stop",
@@ -19510,17 +19946,18 @@
}
},
{
- "id": "mistralai/ministral-3b",
- "name": "Mistral: Ministral 3B",
+ "id": "mistralai/mistral-medium-3.1",
+ "name": "Mistral: Mistral Medium 3.1",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-10-17 02:00:00 +0200",
- "context_window": 32768,
+ "created_at": "2025-08-14 00:33:59 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -19528,22 +19965,24 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.04,
- "output_per_million": 0.04
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 2.0
}
}
},
"metadata": {
- "description": "Ministral 3B is a 3B parameter model optimized for on-device and edge computing. It excels in knowledge, commonsense reasoning, and function-calling, outperforming larger models like Mistral 7B on most benchmarks. Supporting up to 128k context length, it’s ideal for orchestrating agentic workflows and specialist tasks with efficient inference.",
+ "description": "Mistral Medium 3.1 is an updated version of Mistral Medium 3, which is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances state-of-the-art reasoning and multimodal performance with 8× lower cost compared to traditional large models, making it suitable for scalable deployments across professional and industrial use cases.\n\nThe model excels in domains such as coding, STEM reasoning, and enterprise adaptation. It supports hybrid, on-prem, and in-VPC deployments and is optimized for integration into custom workflows. Mistral Medium 3.1 offers competitive accuracy relative to larger models like Claude Sonnet 3.5/3.7, Llama 4 Maverick, and Command R+, while maintaining broad compatibility across cloud environments.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -19552,7 +19991,7 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -19566,17 +20005,19 @@
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "mistralai/ministral-8b",
- "name": "Mistral: Ministral 8B",
+ "id": "mistralai/mistral-nemo",
+ "name": "Mistral: Mistral Nemo",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-10-17 02:00:00 +0200",
- "context_window": 128000,
+ "created_at": "2024-07-19 10:00:00 +1000",
+ "context_window": 32000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -19590,18 +20031,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.09999999999999999
+ "input_per_million": 0.0075,
+ "output_per_million": 0.049999999999999996
}
}
},
"metadata": {
- "description": "Ministral 8B is an 8B parameter model featuring a unique interleaved sliding-window attention pattern for faster, memory-efficient inference. Designed for edge use cases, it supports up to 128k context length and excels in knowledge and reasoning tasks. It outperforms peers in the sub-10B category, making it perfect for low-latency, privacy-first applications.",
+ "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19611,18 +20053,22 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": null
+ "instruct_type": "mistral"
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 32000,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -19630,18 +20076,20 @@
"temperature",
"tool_choice",
"tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-7b-instruct",
- "name": "Mistral: Mistral 7B Instruct",
+ "id": "mistralai/mistral-nemo:free",
+ "name": "Mistral: Mistral Nemo (free)",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-05-27 02:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": 16384,
+ "created_at": "2024-07-19 10:00:00 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -19653,20 +20101,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.028,
- "output_per_million": 0.054
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*",
+ "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19679,8 +20118,8 @@
"instruct_type": "mistral"
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 16384,
+ "context_length": 131072,
+ "max_completion_tokens": 128000,
"is_moderated": false
},
"per_request_limits": null,
@@ -19692,24 +20131,22 @@
"min_p",
"presence_penalty",
"repetition_penalty",
- "response_format",
"seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-7b-instruct-v0.1",
- "name": "Mistral: Mistral 7B Instruct v0.1",
+ "id": "mistralai/mistral-saba",
+ "name": "Mistral: Saba",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2023-09-28 02:00:00 +0200",
- "context_window": 2824,
+ "created_at": "2025-02-18 01:40:39 +1100",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -19723,18 +20160,18 @@
"capabilities": [
"streaming",
"function_calling",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.11,
- "output_per_million": 0.19
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.6
}
}
},
"metadata": {
- "description": "A 7.3B parameter model that outperforms Llama 2 13B on all benchmarks, with optimizations for speed and context length.",
+ "description": "Mistral Saba is a 24B-parameter language model specifically designed for the Middle East and South Asia, delivering accurate and contextually relevant responses while maintaining efficient performance. Trained on curated regional datasets, it supports multiple Indian-origin languages—including Tamil and Malayalam—alongside Arabic. This makes it a versatile option for a range of regional and multilingual applications. Read more at the blog post [here](https://mistral.ai/en/news/mistral-saba)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19744,39 +20181,37 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": "mistral"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 2824,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
- "top_k",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-7b-instruct-v0.3",
- "name": "Mistral: Mistral 7B Instruct v0.3",
+ "id": "mistralai/mistral-small",
+ "name": "Mistral Small",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-05-27 02:00:00 +0200",
+ "created_at": "2024-01-10 11:00:00 +1100",
"context_window": 32768,
- "max_output_tokens": 16384,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -19789,19 +20224,18 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.028,
- "output_per_million": 0.054
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.6
}
}
},
"metadata": {
- "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct v0.2](/models/mistralai/mistral-7b-instruct-v0.2), with the following changes:\n\n- Extended vocabulary to 32768\n- Supports v3 Tokenizer\n- Supports function calling\n\nNOTE: Support for function calling depends on the provider.",
+ "description": "With 22 billion parameters, Mistral Small v24.09 offers a convenient mid-point between (Mistral NeMo 12B)[/mistralai/mistral-nemo] and (Mistral Large 2)[/mistralai/mistral-large], providing a cost-effective solution that can be deployed across various platforms and environments. It has better reasoning, exhibits more capabilities, can produce and reason about code, and is multiligual, supporting English, French, German, Italian, and Spanish.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19811,41 +20245,37 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": "mistral"
+ "instruct_type": null
},
"top_provider": {
"context_length": 32768,
- "max_completion_tokens": 16384,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
- "top_k",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-7b-instruct:free",
- "name": "Mistral: Mistral 7B Instruct (free)",
+ "id": "mistralai/mistral-small-24b-instruct-2501",
+ "name": "Mistral: Mistral Small 3",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-05-27 02:00:00 +0200",
+ "created_at": "2025-01-31 03:43:29 +1100",
"context_window": 32768,
- "max_output_tokens": 16384,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -19858,11 +20288,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.01999188,
+ "output_per_million": 0.0800064
+ }
+ }
+ },
"metadata": {
- "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*",
+ "description": "Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed for efficient local deployment.\n\nThe model achieves 81% accuracy on the MMLU benchmark and performs competitively with larger models like Llama 3.3 70B and Qwen 32B, while operating at three times the speed on equivalent hardware. [Read the blog post about the model here.](https://mistral.ai/news/mistral-small-3/)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19872,16 +20310,18 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": "mistral"
+ "instruct_type": null
},
"top_provider": {
"context_length": 32768,
- "max_completion_tokens": 16384,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -19889,21 +20329,23 @@
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-large",
- "name": "Mistral Large",
+ "id": "mistralai/mistral-small-24b-instruct-2501:free",
+ "name": "Mistral: Mistral Small 3 (free)",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-02-26 01:00:00 +0100",
- "context_window": 128000,
+ "created_at": "2025-01-31 03:43:29 +1100",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -19916,19 +20358,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 2.0,
- "output_per_million": 6.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "This is Mistral AI's flagship model, Mistral Large 2 (version `mistral-large-2407`). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.",
+ "description": "Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed for efficient local deployment.\n\nThe model achieves 81% accuracy on the MMLU benchmark and performs competitively with larger models like Llama 3.3 70B and Qwen 32B, while operating at three times the speed on equivalent hardware. [Read the blog post about the model here.](https://mistral.ai/news/mistral-small-3/)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -19941,38 +20375,41 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-large-2407",
- "name": "Mistral Large 2407",
+ "id": "mistralai/mistral-small-3.1-24b-instruct",
+ "name": "Mistral: Mistral Small 3.1 24B",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-11-19 02:06:55 +0100",
+ "created_at": "2025-03-18 06:15:37 +1100",
"context_window": 131072,
- "max_output_tokens": null,
+ "max_output_tokens": 96000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -19981,22 +20418,24 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 6.0
+ "input_per_million": 0.01999188,
+ "output_per_million": 0.0800064
}
}
},
"metadata": {
- "description": "This is Mistral AI's flagship model, Mistral Large 2 (version mistral-large-2407). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.\n",
+ "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments. The updated version is [Mistral Small 3.2](mistralai/mistral-small-3.2-24b-instruct)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -20006,14 +20445,18 @@
},
"top_provider": {
"context_length": 131072,
- "max_completion_tokens": null,
+ "max_completion_tokens": 96000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -20021,22 +20464,25 @@
"temperature",
"tool_choice",
"tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-large-2411",
- "name": "Mistral Large 2411",
+ "id": "mistralai/mistral-small-3.1-24b-instruct:free",
+ "name": "Mistral: Mistral Small 3.1 24B (free)",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-11-19 02:11:25 +0100",
- "context_window": 131072,
+ "created_at": "2025-03-18 06:15:37 +1100",
+ "context_window": 128000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -20045,22 +20491,17 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 2.0,
- "output_per_million": 6.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Mistral Large 2 2411 is an update of [Mistral Large 2](/mistralai/mistral-large) released together with [Pixtral Large 2411](/mistralai/pixtral-large-2411)\n\nIt provides a significant upgrade on the previous [Mistral Large 24.07](/mistralai/mistral-large-2407), with notable improvements in long context understanding, a new system prompt, and more accurate function calling.",
+ "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments. The updated version is [Mistral Small 3.2](mistralai/mistral-small-3.2-24b-instruct)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -20069,15 +20510,19 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 128000,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -20085,23 +20530,25 @@
"temperature",
"tool_choice",
"tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-medium-3",
- "name": "Mistral: Mistral Medium 3",
+ "id": "mistralai/mistral-small-3.2-24b-instruct",
+ "name": "Mistral: Mistral Small 3.2 24B",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-05-07 16:15:41 +0200",
- "context_window": 131072,
+ "created_at": "2025-06-21 04:10:16 +1000",
+ "context_window": 128000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "image",
+ "text"
],
"output": [
"text"
@@ -20110,23 +20557,24 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 2.0
+ "input_per_million": 0.049999999999999996,
+ "output_per_million": 0.09999999999999999
}
}
},
"metadata": {
- "description": "Mistral Medium 3 is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances state-of-the-art reasoning and multimodal performance with 8× lower cost compared to traditional large models, making it suitable for scalable deployments across professional and industrial use cases.\n\nThe model excels in domains such as coding, STEM reasoning, and enterprise adaptation. It supports hybrid, on-prem, and in-VPC deployments and is optimized for integration into custom workflows. Mistral Medium 3 offers competitive accuracy relative to larger models like Claude Sonnet 3.5/3.7, Llama 4 Maverick, and Command R+, while maintaining broad compatibility across cloud environments.",
+ "description": "Mistral-Small-3.2-24B-Instruct-2506 is an updated 24B parameter model from Mistral optimized for instruction following, repetition reduction, and improved function calling. Compared to the 3.1 release, version 3.2 significantly improves accuracy on WildBench and Arena Hard, reduces infinite generations, and delivers gains in tool use and structured output tasks.\n\nIt supports image and text inputs with structured outputs, function/tool calling, and strong performance across coding (HumanEval+, MBPP), STEM (MMLU, MATH, GPQA), and vision benchmarks (ChartQA, DocVQA).",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "text",
- "image"
+ "image",
+ "text"
],
"output_modalities": [
"text"
@@ -20135,15 +20583,19 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 128000,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -20151,23 +20603,25 @@
"temperature",
"tool_choice",
"tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-medium-3.1",
- "name": "Mistral: Mistral Medium 3.1",
+ "id": "mistralai/mistral-small-3.2-24b-instruct:free",
+ "name": "Mistral: Mistral Small 3.2 24B (free)",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-08-13 16:33:59 +0200",
- "context_window": 262144,
+ "created_at": "2025-06-21 04:10:16 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "image",
+ "text"
],
"output": [
"text"
@@ -20176,23 +20630,16 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 2.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Mistral Medium 3.1 is an updated version of Mistral Medium 3, which is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances state-of-the-art reasoning and multimodal performance with 8× lower cost compared to traditional large models, making it suitable for scalable deployments across professional and industrial use cases.\n\nThe model excels in domains such as coding, STEM reasoning, and enterprise adaptation. It supports hybrid, on-prem, and in-VPC deployments and is optimized for integration into custom workflows. Mistral Medium 3.1 offers competitive accuracy relative to larger models like Claude Sonnet 3.5/3.7, Llama 4 Maverick, and Command R+, while maintaining broad compatibility across cloud environments.",
+ "description": "Mistral-Small-3.2-24B-Instruct-2506 is an updated 24B parameter model from Mistral optimized for instruction following, repetition reduction, and improved function calling. Compared to the 3.1 release, version 3.2 significantly improves accuracy on WildBench and Arena Hard, reduces infinite generations, and delivers gains in tool use and structured output tasks.\n\nIt supports image and text inputs with structured outputs, function/tool calling, and strong performance across coding (HumanEval+, MBPP), STEM (MMLU, MATH, GPQA), and vision benchmarks (ChartQA, DocVQA).",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "text",
- "image"
+ "image",
+ "text"
],
"output_modalities": [
"text"
@@ -20201,33 +20648,38 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 262144,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "repetition_penalty",
"seed",
"stop",
"structured_outputs",
"temperature",
"tool_choice",
"tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-nemo",
- "name": "Mistral: Mistral Nemo",
+ "id": "mistralai/mistral-tiny",
+ "name": "Mistral Tiny",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-07-19 02:00:00 +0200",
- "context_window": 32000,
+ "created_at": "2024-01-10 11:00:00 +1100",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -20241,19 +20693,18 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.0075,
- "output_per_million": 0.049999999999999996
+ "input_per_million": 0.25,
+ "output_per_million": 0.25
}
}
},
"metadata": {
- "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.",
+ "description": "Note: This model is being deprecated. Recommended replacement is the newer [Ministral 8B](/mistral/ministral-8b)\n\nThis model is currently powered by Mistral-7B-v0.2, and incorporates a \"better\" fine-tuning than [Mistral 7B](/models/mistralai/mistral-7b-instruct-v0.1), inspired by community work. It's best used for large batch processing tasks where cost is a significant factor but reasoning capabilities are not crucial.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -20263,22 +20714,18 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": "mistral"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32000,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -20286,20 +20733,18 @@
"temperature",
"tool_choice",
"tools",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-nemo:free",
- "name": "Mistral: Mistral Nemo (free)",
+ "id": "mistralai/mixtral-8x22b-instruct",
+ "name": "Mistral: Mixtral 8x22B Instruct",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-07-19 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": 128000,
+ "created_at": "2024-04-17 10:00:00 +1000",
+ "context_window": 65536,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -20311,11 +20756,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.8999999999999999,
+ "output_per_million": 0.8999999999999999
+ }
+ }
+ },
"metadata": {
- "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.",
+ "description": "Mistral's official instruct fine-tuned version of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b). It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include:\n- strong math, coding, and reasoning\n- large context length (64k)\n- fluency in English, French, Italian, German, and Spanish\n\nSee benchmarks on the launch announcement [here](https://mistral.ai/news/mixtral-8x22b/).\n#moe",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -20328,8 +20782,8 @@
"instruct_type": "mistral"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 128000,
+ "context_length": 65536,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -20338,12 +20792,15 @@
"logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -20351,13 +20808,13 @@
}
},
{
- "id": "mistralai/mistral-saba",
- "name": "Mistral: Saba",
+ "id": "mistralai/mixtral-8x7b-instruct",
+ "name": "Mistral: Mixtral 8x7B Instruct",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-02-17 15:40:39 +0100",
+ "created_at": "2023-12-10 11:00:00 +1100",
"context_window": 32768,
- "max_output_tokens": null,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -20370,18 +20827,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.6
+ "input_per_million": 0.08,
+ "output_per_million": 0.24
}
}
},
"metadata": {
- "description": "Mistral Saba is a 24B-parameter language model specifically designed for the Middle East and South Asia, delivering accurate and contextually relevant responses while maintaining efficient performance. Trained on curated regional datasets, it supports multiple Indian-origin languages—including Tamil and Malayalam—alongside Arabic. This makes it a versatile option for a range of regional and multilingual applications. Read more at the blog post [here](https://mistral.ai/en/news/mistral-saba)",
+ "description": "Mixtral 8x7B Instruct is a pretrained generative Sparse Mixture of Experts, by Mistral AI, for chat and instruction use. Incorporates 8 experts (feed-forward networks) for a total of 47 billion parameters.\n\nInstruct model fine-tuned by Mistral. #moe",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -20391,41 +20849,45 @@
"text"
],
"tokenizer": "Mistral",
- "instruct_type": null
+ "instruct_type": "mistral"
},
"top_provider": {
"context_length": 32768,
- "max_completion_tokens": null,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
+ "top_k",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-small",
- "name": "Mistral Small",
+ "id": "mistralai/pixtral-12b",
+ "name": "Mistral: Pixtral 12B",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2024-01-10 01:00:00 +0100",
+ "created_at": "2024-09-10 10:00:00 +1000",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -20434,22 +20896,24 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.6
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.09999999999999999
}
}
},
"metadata": {
- "description": "With 22 billion parameters, Mistral Small v24.09 offers a convenient mid-point between (Mistral NeMo 12B)[/mistralai/mistral-nemo] and (Mistral Large 2)[/mistralai/mistral-large], providing a cost-effective solution that can be deployed across various platforms and environments. It has better reasoning, exhibits more capabilities, can produce and reason about code, and is multiligual, supporting English, French, German, Italian, and Spanish.",
+ "description": "The first multi-modal, text+image-to-text model from Mistral AI. Its weights were launched via torrent: https://x.com/mistralai/status/1833758285167722836.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -20465,8 +20929,12 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -20474,22 +20942,25 @@
"temperature",
"tool_choice",
"tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-small-24b-instruct-2501",
- "name": "Mistral: Mistral Small 3",
+ "id": "mistralai/pixtral-large-2411",
+ "name": "Mistral: Pixtral Large 2411",
"provider": "openrouter",
"family": "mistralai",
- "created_at": "2025-01-30 17:43:29 +0100",
- "context_window": 32768,
+ "created_at": "2024-11-19 11:49:48 +1100",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -20498,23 +20969,23 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01999188,
- "output_per_million": 0.0800064
+ "input_per_million": 2.0,
+ "output_per_million": 6.0
}
}
},
"metadata": {
- "description": "Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed for efficient local deployment.\n\nThe model achieves 81% accuracy on the MMLU benchmark and performs competitively with larger models like Llama 3.3 70B and Qwen 32B, while operating at three times the speed on equivalent hardware. [Read the blog post about the model here.](https://mistral.ai/news/mistral-small-3/)",
+ "description": "Pixtral Large is a 124B parameter, open-weight, multimodal model built on top of [Mistral Large 2](/mistralai/mistral-large-2411). The model is able to understand documents, charts and natural images.\n\nThe model is available under the Mistral Research License (MRL) for research and educational use, and the Mistral Commercial License for experimentation, testing, and production for commercial purposes.\n\n",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -20523,19 +20994,15 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -20543,19 +21010,17 @@
"temperature",
"tool_choice",
"tools",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mistral-small-24b-instruct-2501:free",
- "name": "Mistral: Mistral Small 3 (free)",
+ "id": "moonshotai/kimi-dev-72b:free",
+ "name": "MoonshotAI: Kimi Dev 72B (free)",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2025-01-30 17:43:29 +0100",
- "context_window": 32768,
+ "family": "moonshotai",
+ "created_at": "2025-06-17 09:18:29 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -20572,7 +21037,7 @@
],
"pricing": {},
"metadata": {
- "description": "Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed for efficient local deployment.\n\nThe model achieves 81% accuracy on the MMLU benchmark and performs competitively with larger models like Llama 3.3 70B and Qwen 32B, while operating at three times the speed on equivalent hardware. [Read the blog post about the model here.](https://mistral.ai/news/mistral-small-3/)",
+ "description": "Kimi-Dev-72B is an open-source large language model fine-tuned for software engineering and issue resolution tasks. Based on Qwen2.5-72B, it is optimized using large-scale reinforcement learning that applies code patches in real repositories and validates them via full test suite execution—rewarding only correct, robust completions. The model achieves 60.4% on SWE-bench Verified, setting a new benchmark among open-source models for software bug fixing and code reasoning.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -20581,22 +21046,24 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
"seed",
"stop",
@@ -20608,18 +21075,17 @@
}
},
{
- "id": "mistralai/mistral-small-3.1-24b-instruct",
- "name": "Mistral: Mistral Small 3.1 24B",
+ "id": "moonshotai/kimi-k2",
+ "name": "MoonshotAI: Kimi K2",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2025-03-17 20:15:37 +0100",
- "context_window": 131072,
- "max_output_tokens": 96000,
+ "family": "moonshotai",
+ "created_at": "2025-07-12 05:47:32 +1000",
+ "context_window": 63000,
+ "max_output_tokens": 63000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -20634,28 +21100,27 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01999188,
- "output_per_million": 0.0800064
+ "input_per_million": 0.14,
+ "output_per_million": 2.4899999999999998
}
}
},
"metadata": {
- "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments. The updated version is [Mistral Small 3.2](mistralai/mistral-small-3.2-24b-instruct)",
+ "description": "Kimi K2 Instruct is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. It is optimized for agentic capabilities, including advanced tool use, reasoning, and code synthesis. Kimi K2 excels across a broad range of benchmarks, particularly in coding (LiveCodeBench, SWE-bench), reasoning (ZebraLogic, GPQA), and tool-use (Tau2, AceBench) tasks. It supports long-context inference up to 128K tokens and is designed with a novel training stack that includes the MuonClip optimizer for stable large-scale MoE training.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 96000,
+ "context_length": 63000,
+ "max_completion_tokens": 63000,
"is_moderated": false
},
"per_request_limits": null,
@@ -20681,18 +21146,17 @@
}
},
{
- "id": "mistralai/mistral-small-3.1-24b-instruct:free",
- "name": "Mistral: Mistral Small 3.1 24B (free)",
+ "id": "moonshotai/kimi-k2:free",
+ "name": "MoonshotAI: Kimi K2 (free)",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2025-03-17 20:15:37 +0100",
- "context_window": 128000,
+ "family": "moonshotai",
+ "created_at": "2025-07-12 05:47:32 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -20701,26 +21165,24 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
"predicted_outputs"
],
"pricing": {},
"metadata": {
- "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments. The updated version is [Mistral Small 3.2](mistralai/mistral-small-3.2-24b-instruct)",
+ "description": "Kimi K2 Instruct is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. It is optimized for agentic capabilities, including advanced tool use, reasoning, and code synthesis. Kimi K2 excels across a broad range of benchmarks, particularly in coding (LiveCodeBench, SWE-bench), reasoning (ZebraLogic, GPQA), and tool-use (Tau2, AceBench) tasks. It supports long-context inference up to 128K tokens and is designed with a novel training stack that includes the MuonClip optimizer for stable large-scale MoE training.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -20733,10 +21195,8 @@
"min_p",
"presence_penalty",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -20747,12 +21207,12 @@
}
},
{
- "id": "mistralai/mistral-small-3.2-24b-instruct",
- "name": "Mistral: Mistral Small 3.2 24B",
+ "id": "moonshotai/kimi-vl-a3b-thinking",
+ "name": "MoonshotAI: Kimi VL A3B Thinking",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2025-06-20 20:10:16 +0200",
- "context_window": 128000,
+ "family": "moonshotai",
+ "created_at": "2025-04-11 03:07:21 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -20766,20 +21226,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.049999999999999996,
- "output_per_million": 0.09999999999999999
+ "input_per_million": 0.02498985,
+ "output_per_million": 0.100008
}
}
},
"metadata": {
- "description": "Mistral-Small-3.2-24B-Instruct-2506 is an updated 24B parameter model from Mistral optimized for instruction following, repetition reduction, and improved function calling. Compared to the 3.1 release, version 3.2 significantly improves accuracy on WildBench and Arena Hard, reduces infinite generations, and delivers gains in tool use and structured output tasks.\n\nIt supports image and text inputs with structured outputs, function/tool calling, and strong performance across coding (HumanEval+, MBPP), STEM (MMLU, MATH, GPQA), and vision benchmarks (ChartQA, DocVQA).",
+ "description": "Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -20789,30 +21247,28 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -20820,11 +21276,11 @@
}
},
{
- "id": "mistralai/mistral-small-3.2-24b-instruct:free",
- "name": "Mistral: Mistral Small 3.2 24B (free)",
+ "id": "moonshotai/kimi-vl-a3b-thinking:free",
+ "name": "MoonshotAI: Kimi VL A3B Thinking (free)",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2025-06-20 20:10:16 +0200",
+ "family": "moonshotai",
+ "created_at": "2025-04-11 03:07:21 +1000",
"context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -20839,12 +21295,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
"predicted_outputs"
],
"pricing": {},
"metadata": {
- "description": "Mistral-Small-3.2-24B-Instruct-2506 is an updated 24B parameter model from Mistral optimized for instruction following, repetition reduction, and improved function calling. Compared to the 3.1 release, version 3.2 significantly improves accuracy on WildBench and Arena Hard, reduces infinite generations, and delivers gains in tool use and structured output tasks.\n\nIt supports image and text inputs with structured outputs, function/tool calling, and strong performance across coding (HumanEval+, MBPP), STEM (MMLU, MATH, GPQA), and vision benchmarks (ChartQA, DocVQA).",
+ "description": "Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -20854,7 +21309,7 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
@@ -20865,18 +21320,17 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -20884,13 +21338,13 @@
}
},
{
- "id": "mistralai/mistral-tiny",
- "name": "Mistral Tiny",
+ "id": "morph/morph-v3-fast",
+ "name": "Morph: Morph V3 Fast",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2024-01-10 01:00:00 +0100",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "morph",
+ "created_at": "2025-07-08 03:40:02 +1000",
+ "context_window": 81920,
+ "max_output_tokens": 38000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -20901,20 +21355,18 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "output_per_million": 0.25
+ "input_per_million": 0.8999999999999999,
+ "output_per_million": 1.9
}
}
},
"metadata": {
- "description": "Note: This model is being deprecated. Recommended replacement is the newer [Ministral 8B](/mistral/ministral-8b)\n\nThis model is currently powered by Mistral-7B-v0.2, and incorporates a \"better\" fine-tuning than [Mistral 7B](/models/mistralai/mistral-7b-instruct-v0.1), inspired by community work. It's best used for large batch processing tasks where cost is a significant factor but reasoning capabilities are not crucial.",
+ "description": "Morph's fastest apply model for code edits. 4500+ tokens/sec with 96% accuracy for rapid code transformations.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -20923,38 +21375,30 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 81920,
+ "max_completion_tokens": 38000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"max_tokens",
- "presence_penalty",
- "response_format",
- "seed",
"stop",
- "structured_outputs",
- "temperature",
- "tool_choice",
- "tools",
- "top_p"
+ "temperature"
]
}
},
{
- "id": "mistralai/mixtral-8x22b-instruct",
- "name": "Mistral: Mixtral 8x22B Instruct",
+ "id": "morph/morph-v3-large",
+ "name": "Morph: Morph V3 Large",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2024-04-17 02:00:00 +0200",
- "context_window": 65536,
- "max_output_tokens": null,
+ "family": "morph",
+ "created_at": "2025-07-08 03:54:18 +1000",
+ "context_window": 81920,
+ "max_output_tokens": 38000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -20965,21 +21409,18 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "predicted_outputs"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.8999999999999999,
- "output_per_million": 0.8999999999999999
+ "output_per_million": 1.9
}
}
},
"metadata": {
- "description": "Mistral's official instruct fine-tuned version of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b). It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include:\n- strong math, coding, and reasoning\n- large context length (64k)\n- fluency in English, French, Italian, German, and Spanish\n\nSee benchmarks on the launch announcement [here](https://mistral.ai/news/mixtral-8x22b/).\n#moe",
+ "description": "Morph's high-accuracy apply model for complex code edits. 2000+ tokens/sec with 98% accuracy for precise code transformations.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -20988,43 +21429,91 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
- "instruct_type": "mistral"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 65536,
- "max_completion_tokens": null,
+ "context_length": 81920,
+ "max_completion_tokens": 38000,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "max_tokens",
+ "stop",
+ "temperature"
+ ]
+ }
+ },
+ {
+ "id": "neversleep/llama-3-lumimaid-70b",
+ "name": "NeverSleep: Llama 3 Lumimaid 70B",
+ "provider": "openrouter",
+ "family": "neversleep",
+ "created_at": "2024-05-16 10:00:00 +1000",
+ "context_window": 8192,
+ "max_output_tokens": 4096,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 4.0,
+ "output_per_million": 6.0
+ }
+ }
+ },
+ "metadata": {
+ "description": "The NeverSleep team is back, with a Llama 3 70B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
+ },
+ "top_provider": {
+ "context_length": 8192,
+ "max_completion_tokens": 4096,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/mixtral-8x7b-instruct",
- "name": "Mistral: Mixtral 8x7B Instruct",
+ "id": "neversleep/llama-3.1-lumimaid-8b",
+ "name": "NeverSleep: Lumimaid v0.2 8B",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2023-12-10 01:00:00 +0100",
+ "family": "neversleep",
+ "created_at": "2024-09-15 10:00:00 +1000",
"context_window": 32768,
- "max_output_tokens": 16384,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -21036,20 +21525,19 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.08,
- "output_per_million": 0.24
+ "input_per_million": 0.09,
+ "output_per_million": 0.6
}
}
},
"metadata": {
- "description": "Mixtral 8x7B Instruct is a pretrained generative Sparse Mixture of Experts, by Mistral AI, for chat and instruction use. Incorporates 8 experts (feed-forward networks) for a total of 47 billion parameters.\n\nInstruct model fine-tuned by Mistral. #moe",
+ "description": "Lumimaid v0.2 8B is a finetune of [Llama 3.1 8B](/models/meta-llama/llama-3.1-8b-instruct) with a \"HUGE step up dataset wise\" compared to Lumimaid v0.1. Sloppy chats output were purged.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21058,12 +21546,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
- "instruct_type": "mistral"
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
"context_length": 32768,
- "max_completion_tokens": 16384,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -21077,27 +21565,26 @@
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_a",
"top_k",
"top_p"
]
}
},
{
- "id": "mistralai/pixtral-12b",
- "name": "Mistral: Pixtral 12B",
+ "id": "neversleep/noromaid-20b",
+ "name": "Noromaid 20B",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2024-09-10 02:00:00 +0200",
- "context_window": 32768,
+ "family": "neversleep",
+ "created_at": "2023-11-26 11:00:00 +1100",
+ "context_window": 4096,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -21105,34 +21592,32 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.09999999999999999
+ "input_per_million": 1.0,
+ "output_per_million": 1.75
}
}
},
"metadata": {
- "description": "The first multi-modal, text+image-to-text model from Mistral AI. Its weights were launched via torrent: https://x.com/mistralai/status/1833758285167722836.",
+ "description": "A collab between IkariDev and Undi. This merge is suitable for RP, ERP, and general knowledge.\n\n#merge #uncensored",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
- "instruct_type": null
+ "tokenizer": "Llama2",
+ "instruct_type": "alpaca"
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 4096,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -21140,7 +21625,6 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -21150,27 +21634,24 @@
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_a",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "mistralai/pixtral-large-2411",
- "name": "Mistral: Pixtral Large 2411",
+ "id": "nousresearch/deephermes-3-llama-3-8b-preview:free",
+ "name": "Nous: DeepHermes 3 Llama 3 8B Preview (free)",
"provider": "openrouter",
- "family": "mistralai",
- "created_at": "2024-11-19 01:49:48 +0100",
+ "family": "nousresearch",
+ "created_at": "2025-02-28 16:09:32 +1100",
"context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [
- "text",
- "image"
+ "input": [
+ "text"
],
"output": [
"text"
@@ -21178,29 +21659,20 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 2.0,
- "output_per_million": 6.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Pixtral Large is a 124B parameter, open-weight, multimodal model built on top of [Mistral Large 2](/mistralai/mistral-large-2411). The model is able to understand documents, charts and natural images.\n\nThe model is available under the Mistral Research License (MRL) for research and educational use, and the Mistral Commercial License for experimentation, testing, and production for commercial purposes.\n\n",
+ "description": "DeepHermes 3 Preview is the latest version of our flagship Hermes series of LLMs by Nous Research, and one of the first models in the world to unify Reasoning (long chains of thought that improve answer accuracy) and normal LLM response modes into one model. We have also improved LLM annotation, judgement, and function calling.\n\nDeepHermes 3 Preview is one of the first LLM models to unify both \"intuitive\", traditional mode responses and long chain of thought reasoning responses into a single model, toggled by a system prompt.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
@@ -21211,26 +21683,28 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "moonshotai/kimi-dev-72b:free",
- "name": "MoonshotAI: Kimi Dev 72B (free)",
+ "id": "nousresearch/deephermes-3-mistral-24b-preview",
+ "name": "Nous: DeepHermes 3 Mistral 24B Preview",
"provider": "openrouter",
- "family": "moonshotai",
- "created_at": "2025-06-17 01:18:29 +0200",
- "context_window": 131072,
+ "family": "nousresearch",
+ "created_at": "2025-05-10 08:48:24 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -21245,9 +21719,16 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.09329544,
+ "output_per_million": 0.3733632
+ }
+ }
+ },
"metadata": {
- "description": "Kimi-Dev-72B is an open-source large language model fine-tuned for software engineering and issue resolution tasks. Based on Qwen2.5-72B, it is optimized using large-scale reinforcement learning that applies code patches in real repositories and validates them via full test suite execution—rewarding only correct, robust completions. The model achieves 60.4% on SWE-bench Verified, setting a new benchmark among open-source models for software bug fixing and code reasoning.",
+ "description": "DeepHermes 3 (Mistral 24B Preview) is an instruction-tuned language model by Nous Research based on Mistral-Small-24B, designed for chat, function calling, and advanced multi-turn reasoning. It introduces a dual-mode system that toggles between intuitive chat responses and structured “deep reasoning” mode using special system prompts. Fine-tuned via distillation from R1, it supports structured output (JSON mode) and function call syntax for agent-based applications.\n\nDeepHermes 3 supports a **reasoning toggle via system prompt**, allowing users to switch between fast, intuitive responses and deliberate, multi-step reasoning. When activated with the following specific system instruction, the model enters a *\"deep thinking\"* mode—generating extended chains of thought wrapped in `` tags before delivering a final answer. \n\nSystem Prompt: You are a deep thinking AI, you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself via systematic reasoning processes to help come to a correct solution prior to answering. You should enclose your thoughts and internal monologue inside tags, and then provide your solution or response to the problem.\n",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21260,7 +21741,7 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -21285,13 +21766,13 @@
}
},
{
- "id": "moonshotai/kimi-k2",
- "name": "MoonshotAI: Kimi K2",
+ "id": "nousresearch/hermes-2-pro-llama-3-8b",
+ "name": "NousResearch: Hermes 2 Pro - Llama-3 8B",
"provider": "openrouter",
- "family": "moonshotai",
- "created_at": "2025-07-11 21:47:32 +0200",
- "context_window": 63000,
- "max_output_tokens": 63000,
+ "family": "nousresearch",
+ "created_at": "2024-05-27 10:00:00 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 131072,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -21303,20 +21784,19 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.14,
- "output_per_million": 2.4899999999999998
+ "input_per_million": 0.024999999999999998,
+ "output_per_million": 0.04
}
}
},
"metadata": {
- "description": "Kimi K2 Instruct is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. It is optimized for agentic capabilities, including advanced tool use, reasoning, and code synthesis. Kimi K2 excels across a broad range of benchmarks, particularly in coding (LiveCodeBench, SWE-bench), reasoning (ZebraLogic, GPQA), and tool-use (Tau2, AceBench) tasks. It supports long-context inference up to 128K tokens and is designed with a novel training stack that includes the MuonClip optimizer for stable large-scale MoE training.",
+ "description": "Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21325,12 +21805,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 63000,
- "max_completion_tokens": 63000,
+ "context_length": 131072,
+ "max_completion_tokens": 131072,
"is_moderated": false
},
"per_request_limits": null,
@@ -21347,8 +21827,6 @@
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -21356,13 +21834,13 @@
}
},
{
- "id": "moonshotai/kimi-k2:free",
- "name": "MoonshotAI: Kimi K2 (free)",
+ "id": "nousresearch/hermes-3-llama-3.1-405b",
+ "name": "Nous: Hermes 3 405B Instruct",
"provider": "openrouter",
- "family": "moonshotai",
- "created_at": "2025-07-11 21:47:32 +0200",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "nousresearch",
+ "created_at": "2024-08-16 10:00:00 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -21374,12 +21852,19 @@
},
"capabilities": [
"streaming",
- "function_calling",
+ "structured_output",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.7,
+ "output_per_million": 0.7999999999999999
+ }
+ }
+ },
"metadata": {
- "description": "Kimi K2 Instruct is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. It is optimized for agentic capabilities, including advanced tool use, reasoning, and code synthesis. Kimi K2 excels across a broad range of benchmarks, particularly in coding (LiveCodeBench, SWE-bench), reasoning (ZebraLogic, GPQA), and tool-use (Tau2, AceBench) tasks. It supports long-context inference up to 128K tokens and is designed with a novel training stack that includes the MuonClip optimizer for stable large-scale MoE training.",
+ "description": "Hermes 3 is a generalist language model with many improvements over Hermes 2, including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 405B is a frontier-level, full-parameter finetune of the Llama-3.1 405B foundation model, focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.\n\nHermes 3 is competitive, if not superior, to Llama-3.1 Instruct models at general capabilities, with varying strengths and weaknesses attributable between the two.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21388,12 +21873,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
@@ -21405,11 +21890,10 @@
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -21417,17 +21901,16 @@
}
},
{
- "id": "moonshotai/kimi-vl-a3b-thinking",
- "name": "MoonshotAI: Kimi VL A3B Thinking",
+ "id": "nousresearch/hermes-3-llama-3.1-70b",
+ "name": "Nous: Hermes 3 70B Instruct",
"provider": "openrouter",
- "family": "moonshotai",
- "created_at": "2025-04-10 19:07:21 +0200",
+ "family": "nousresearch",
+ "created_at": "2024-08-18 10:00:00 +1000",
"context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -21436,29 +21919,30 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.02498985,
- "output_per_million": 0.100008
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.28
}
}
},
"metadata": {
- "description": "Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.",
+ "description": "Hermes 3 is a generalist language model with many improvements over [Hermes 2](/models/nousresearch/nous-hermes-2-mistral-7b-dpo), including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 70B is a competitive, if not superior finetune of the [Llama-3.1 70B foundation model](/models/meta-llama/llama-3.1-70b-instruct), focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "chatml"
},
"top_provider": {
"context_length": 131072,
@@ -21468,17 +21952,19 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -21486,17 +21972,16 @@
}
},
{
- "id": "moonshotai/kimi-vl-a3b-thinking:free",
- "name": "MoonshotAI: Kimi VL A3B Thinking (free)",
+ "id": "nousresearch/hermes-4-405b",
+ "name": "Nous: Hermes 4 405B",
"provider": "openrouter",
- "family": "moonshotai",
- "created_at": "2025-04-10 19:07:21 +0200",
+ "family": "nousresearch",
+ "created_at": "2025-08-27 05:11:03 +1000",
"context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -21507,13 +21992,19 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.0,
+ "output_per_million": 3.0
+ }
+ }
+ },
"metadata": {
- "description": "Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.",
+ "description": "Hermes 4 is a large-scale reasoning model built on Meta-Llama-3.1-405B and released by Nous Research. It introduces a hybrid reasoning mode, where the model can choose to deliberate internally with ... traces or respond directly, offering flexibility between speed and depth. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)\n\nThe model is instruction-tuned with an expanded post-training corpus (~60B tokens) emphasizing reasoning traces, improving performance in math, code, STEM, and logical reasoning, while retaining broad assistant utility. It also supports structured outputs, including JSON mode, schema adherence, function calling, and tool use. Hermes 4 is trained for steerability, lower refusal rates, and alignment toward neutral, user-directed behavior.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
@@ -21534,10 +22025,8 @@
"logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
"reasoning",
- "repetition_penalty",
"seed",
"stop",
"temperature",
@@ -21548,13 +22037,13 @@
}
},
{
- "id": "morph/morph-v3-fast",
- "name": "Morph: Morph V3 Fast",
+ "id": "nousresearch/hermes-4-70b",
+ "name": "Nous: Hermes 4 70B",
"provider": "openrouter",
- "family": "morph",
- "created_at": "2025-07-07 19:40:02 +0200",
- "context_window": 81920,
- "max_output_tokens": 38000,
+ "family": "nousresearch",
+ "created_at": "2025-08-27 05:23:02 +1000",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -21565,18 +22054,20 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.8999999999999999,
- "output_per_million": 1.9
+ "input_per_million": 0.09329544,
+ "output_per_million": 0.3733632
}
}
},
"metadata": {
- "description": "Morph's fastest apply model for code edits. 4500+ tokens/sec with 96% accuracy for rapid code transformations.",
+ "description": "Hermes 4 70B is a hybrid reasoning model from Nous Research, built on Meta-Llama-3.1-70B. It introduces the same hybrid mode as the larger 405B release, allowing the model to either respond directly or generate explicit ... reasoning traces before answering. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)\n\nThis 70B variant is trained with the expanded post-training corpus (~60B tokens) emphasizing verified reasoning data, leading to improvements in mathematics, coding, STEM, logic, and structured outputs while maintaining general assistant performance. It supports JSON mode, schema adherence, function calling, and tool use, and is designed for greater steerability with reduced refusal rates.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21585,30 +22076,44 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Llama3",
"instruct_type": null
},
"top_provider": {
- "context_length": 81920,
- "max_completion_tokens": 38000,
+ "context_length": 131072,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "reasoning",
+ "repetition_penalty",
+ "seed",
"stop",
- "temperature"
+ "temperature",
+ "tool_choice",
+ "tools",
+ "top_k",
+ "top_logprobs",
+ "top_p"
]
}
},
{
- "id": "morph/morph-v3-large",
- "name": "Morph: Morph V3 Large",
+ "id": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo",
+ "name": "Nous: Hermes 2 Mixtral 8x7B DPO",
"provider": "openrouter",
- "family": "morph",
- "created_at": "2025-07-07 19:54:18 +0200",
- "context_window": 81920,
- "max_output_tokens": 38000,
+ "family": "nousresearch",
+ "created_at": "2024-01-16 11:00:00 +1100",
+ "context_window": 32768,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -21619,18 +22124,19 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.8999999999999999,
- "output_per_million": 1.9
+ "input_per_million": 0.6,
+ "output_per_million": 0.6
}
}
},
"metadata": {
- "description": "Morph's high-accuracy apply model for complex code edits. 2000+ tokens/sec with 98% accuracy for precise code transformations.",
+ "description": "Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the [Mixtral 8x7B MoE LLM](/models/mistralai/mixtral-8x7b).\n\nThe model was trained on over 1,000,000 entries of primarily [GPT-4](/models/openai/gpt-4) generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks.\n\n#moe",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21639,30 +22145,37 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Mistral",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 81920,
- "max_completion_tokens": 38000,
+ "context_length": 32768,
+ "max_completion_tokens": 2048,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"stop",
- "temperature"
+ "temperature",
+ "top_k",
+ "top_p"
]
}
},
{
- "id": "neversleep/llama-3-lumimaid-70b",
- "name": "NeverSleep: Llama 3 Lumimaid 70B",
+ "id": "nvidia/llama-3.1-nemotron-70b-instruct",
+ "name": "NVIDIA: Llama 3.1 Nemotron 70B Instruct",
"provider": "openrouter",
- "family": "neversleep",
- "created_at": "2024-05-16 02:00:00 +0200",
- "context_window": 8192,
- "max_output_tokens": 4096,
+ "family": "nvidia",
+ "created_at": "2024-10-15 11:00:00 +1100",
+ "context_window": 131072,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -21673,18 +22186,21 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 4.0,
- "output_per_million": 6.0
+ "input_per_million": 0.12,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "description": "The NeverSleep team is back, with a Llama 3 70B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "NVIDIA's Llama 3.1 Nemotron 70B is a language model designed for generating precise and useful responses. Leveraging [Llama 3.1 70B](/models/meta-llama/llama-3.1-70b-instruct) architecture and Reinforcement Learning from Human Feedback (RLHF), it excels in automatic alignment benchmarks. This model is tailored for applications requiring high accuracy in helpfulness and response generation, suitable for diverse user queries across multiple domains.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21697,32 +22213,36 @@
"instruct_type": "llama3"
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 4096,
+ "context_length": 131072,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "neversleep/llama-3.1-lumimaid-8b",
- "name": "NeverSleep: Lumimaid v0.2 8B",
+ "id": "nvidia/llama-3.1-nemotron-ultra-253b-v1",
+ "name": "NVIDIA: Llama 3.1 Nemotron Ultra 253B v1",
"provider": "openrouter",
- "family": "neversleep",
- "created_at": "2024-09-15 02:00:00 +0200",
- "context_window": 32768,
+ "family": "nvidia",
+ "created_at": "2025-04-08 22:24:19 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -21735,19 +22255,18 @@
},
"capabilities": [
"streaming",
- "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09,
- "output_per_million": 0.6
+ "input_per_million": 0.6,
+ "output_per_million": 1.7999999999999998
}
}
},
"metadata": {
- "description": "Lumimaid v0.2 8B is a finetune of [Llama 3.1 8B](/models/meta-llama/llama-3.1-8b-instruct) with a \"HUGE step up dataset wise\" compared to Lumimaid v0.1. Sloppy chats output were purged.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
+ "description": "Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) optimized for advanced reasoning, human-interactive chat, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta’s Llama-3.1-405B-Instruct, it has been significantly customized using Neural Architecture Search (NAS), resulting in enhanced efficiency, reduced memory usage, and improved inference latency. The model supports a context length of up to 128K tokens and can operate efficiently on an 8x NVIDIA H100 node.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21757,39 +22276,38 @@
"text"
],
"tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
+ "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
- "response_format",
+ "reasoning",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "top_a",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "neversleep/noromaid-20b",
- "name": "Noromaid 20B",
+ "id": "nvidia/llama-3.1-nemotron-ultra-253b-v1:free",
+ "name": "NVIDIA: Llama 3.1 Nemotron Ultra 253B v1 (free)",
"provider": "openrouter",
- "family": "neversleep",
- "created_at": "2023-11-26 01:00:00 +0100",
- "context_window": 4096,
+ "family": "nvidia",
+ "created_at": "2025-04-08 22:24:19 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -21802,19 +22320,11 @@
},
"capabilities": [
"streaming",
- "structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 1.0,
- "output_per_million": 1.75
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "A collab between IkariDev and Undi. This merge is suitable for RP, ERP, and general knowledge.\n\n#merge #uncensored",
+ "description": "Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) optimized for advanced reasoning, human-interactive chat, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta’s Llama-3.1-405B-Instruct, it has been significantly customized using Neural Architecture Search (NAS), resulting in enhanced efficiency, reduced memory usage, and improved inference latency. The model supports a context length of up to 128K tokens and can operate efficiently on an 8x NVIDIA H100 node.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21823,11 +22333,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama2",
- "instruct_type": "alpaca"
+ "tokenizer": "Llama3",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 4096,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -21835,27 +22345,26 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "top_a",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "nousresearch/deephermes-3-llama-3-8b-preview:free",
- "name": "Nous: DeepHermes 3 Llama 3 8B Preview (free)",
+ "id": "nvidia/llama-3.3-nemotron-super-49b-v1",
+ "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1",
"provider": "openrouter",
- "family": "nousresearch",
- "created_at": "2025-02-28 06:09:32 +0100",
+ "family": "nvidia",
+ "created_at": "2025-04-08 23:38:14 +1000",
"context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -21871,9 +22380,16 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.13,
+ "output_per_million": 0.39999999999999997
+ }
+ }
+ },
"metadata": {
- "description": "DeepHermes 3 Preview is the latest version of our flagship Hermes series of LLMs by Nous Research, and one of the first models in the world to unify Reasoning (long chains of thought that improve answer accuracy) and normal LLM response modes into one model. We have also improved LLM annotation, judgement, and function calling.\n\nDeepHermes 3 Preview is one of the first LLM models to unify both \"intuitive\", traditional mode responses and long chain of thought reasoning responses into a single model, toggled by a system prompt.",
+ "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -21896,9 +22412,7 @@
"logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"seed",
"stop",
"temperature",
@@ -21909,17 +22423,18 @@
}
},
{
- "id": "nousresearch/deephermes-3-mistral-24b-preview",
- "name": "Nous: DeepHermes 3 Mistral 24B Preview",
+ "id": "openai/chatgpt-4o-latest",
+ "name": "OpenAI: ChatGPT-4o",
"provider": "openrouter",
- "family": "nousresearch",
- "created_at": "2025-05-10 00:48:24 +0200",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "openai",
+ "created_at": "2024-08-14 10:00:00 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -21927,65 +22442,64 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09329544,
- "output_per_million": 0.3733632
+ "input_per_million": 5.0,
+ "output_per_million": 15.0
}
}
},
"metadata": {
- "description": "DeepHermes 3 (Mistral 24B Preview) is an instruction-tuned language model by Nous Research based on Mistral-Small-24B, designed for chat, function calling, and advanced multi-turn reasoning. It introduces a dual-mode system that toggles between intuitive chat responses and structured “deep reasoning” mode using special system prompts. Fine-tuned via distillation from R1, it supports structured output (JSON mode) and function call syntax for agent-based applications.\n\nDeepHermes 3 supports a **reasoning toggle via system prompt**, allowing users to switch between fast, intuitive responses and deliberate, multi-step reasoning. When activated with the following specific system instruction, the model enters a *\"deep thinking\"* mode—generating extended chains of thought wrapped in `` tags before delivering a final answer. \n\nSystem Prompt: You are a deep thinking AI, you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself via systematic reasoning processes to help come to a correct solution prior to answering. You should enclose your thoughts and internal monologue inside tags, and then provide your solution or response to the problem.\n",
+ "description": "OpenAI ChatGPT 4o is continually updated by OpenAI to point to the current version of GPT-4o used by ChatGPT. It therefore differs slightly from the API version of [GPT-4o](/models/openai/gpt-4o) in that it has additional RLHF. It is intended for research and evaluation.\n\nOpenAI notes that this model is not suited for production use-cases as it may be removed or redirected to another model in the future.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "GPT",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 16384,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "nousresearch/hermes-2-pro-llama-3-8b",
- "name": "NousResearch: Hermes 2 Pro - Llama-3 8B",
+ "id": "openai/codex-mini",
+ "name": "OpenAI: Codex Mini",
"provider": "openrouter",
- "family": "nousresearch",
- "created_at": "2024-05-27 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": 131072,
+ "family": "openai",
+ "created_at": "2025-05-17 01:36:01 +1000",
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -21994,63 +22508,58 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.024999999999999998,
- "output_per_million": 0.04
+ "input_per_million": 1.5,
+ "output_per_million": 6.0,
+ "cached_input_per_million": 0.375
}
}
},
"metadata": {
- "description": "Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.",
+ "description": "codex-mini-latest is a fine-tuned version of o4-mini specifically for use in Codex CLI. For direct use in the API, we recommend starting with gpt-4.1.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
+ "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "chatml"
+ "tokenizer": "GPT",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 131072,
- "is_moderated": false
+ "context_length": 200000,
+ "max_completion_tokens": 100000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
+ "include_reasoning",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
+ "reasoning",
"response_format",
"seed",
- "stop",
"structured_outputs",
- "temperature",
- "top_k",
- "top_logprobs",
- "top_p"
+ "tool_choice",
+ "tools"
]
}
},
{
- "id": "nousresearch/hermes-3-llama-3.1-405b",
- "name": "Nous: Hermes 3 405B Instruct",
+ "id": "openai/gpt-3.5-turbo",
+ "name": "OpenAI: GPT-3.5 Turbo",
"provider": "openrouter",
- "family": "nousresearch",
- "created_at": "2024-08-16 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": 16384,
+ "family": "openai",
+ "created_at": "2023-05-28 10:00:00 +1000",
+ "context_window": 16385,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -22062,19 +22571,19 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7,
- "output_per_million": 0.7999999999999999
+ "input_per_million": 0.5,
+ "output_per_million": 1.5
}
}
},
"metadata": {
- "description": "Hermes 3 is a generalist language model with many improvements over Hermes 2, including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 405B is a frontier-level, full-parameter finetune of the Llama-3.1 405B foundation model, focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.\n\nHermes 3 is competitive, if not superior, to Llama-3.1 Instruct models at general capabilities, with varying strengths and weaknesses attributable between the two.",
+ "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -22083,13 +22592,13 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "chatml"
+ "tokenizer": "GPT",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 16384,
- "is_moderated": false
+ "context_length": 16385,
+ "max_completion_tokens": 4096,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
@@ -22097,27 +22606,27 @@
"logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "nousresearch/hermes-3-llama-3.1-70b",
- "name": "Nous: Hermes 3 70B Instruct",
+ "id": "openai/gpt-3.5-turbo-0613",
+ "name": "OpenAI: GPT-3.5 Turbo (older v0613)",
"provider": "openrouter",
- "family": "nousresearch",
- "created_at": "2024-08-18 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "family": "openai",
+ "created_at": "2024-01-25 11:00:00 +1100",
+ "context_window": 4095,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -22130,19 +22639,18 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.28
+ "input_per_million": 1.0,
+ "output_per_million": 2.0
}
}
},
"metadata": {
- "description": "Hermes 3 is a generalist language model with many improvements over [Hermes 2](/models/nousresearch/nous-hermes-2-mistral-7b-dpo), including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.\n\nHermes 3 70B is a competitive, if not superior finetune of the [Llama-3.1 70B foundation model](/models/meta-llama/llama-3.1-70b-instruct), focused on aligning LLMs to the user, with powerful steering capabilities and control given to the end user.\n\nThe Hermes 3 series builds and expands on the Hermes 2 set of capabilities, including more powerful and reliable function calling and structured output capabilities, generalist assistant capabilities, and improved code generation skills.",
+ "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -22151,12 +22659,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "chatml"
+ "tokenizer": "GPT",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
+ "context_length": 4095,
+ "max_completion_tokens": 4096,
"is_moderated": false
},
"per_request_limits": null,
@@ -22165,9 +22673,7 @@
"logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -22175,20 +22681,19 @@
"temperature",
"tool_choice",
"tools",
- "top_k",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo",
- "name": "Nous: Hermes 2 Mixtral 8x7B DPO",
+ "id": "openai/gpt-3.5-turbo-16k",
+ "name": "OpenAI: GPT-3.5 Turbo 16k",
"provider": "openrouter",
- "family": "nousresearch",
- "created_at": "2024-01-16 01:00:00 +0100",
- "context_window": 32768,
- "max_output_tokens": 2048,
+ "family": "openai",
+ "created_at": "2023-08-28 10:00:00 +1000",
+ "context_window": 16385,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -22200,18 +22705,19 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.6,
- "output_per_million": 0.6
+ "input_per_million": 3.0,
+ "output_per_million": 4.0
}
}
},
"metadata": {
- "description": "Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the [Mixtral 8x7B MoE LLM](/models/mistralai/mixtral-8x7b).\n\nThe model was trained on over 1,000,000 entries of primarily [GPT-4](/models/openai/gpt-4) generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks.\n\n#moe",
+ "description": "This model offers four times the context length of gpt-3.5-turbo, allowing it to support approximately 20 pages of text in a single request at a higher cost. Training data: up to Sep 2021.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -22220,37 +22726,41 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
- "instruct_type": "chatml"
+ "tokenizer": "GPT",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 2048,
- "is_moderated": false
+ "context_length": 16385,
+ "max_completion_tokens": 4096,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
+ "seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "nvidia/llama-3.1-nemotron-70b-instruct",
- "name": "NVIDIA: Llama 3.1 Nemotron 70B Instruct",
+ "id": "openai/gpt-3.5-turbo-instruct",
+ "name": "OpenAI: GPT-3.5 Turbo Instruct",
"provider": "openrouter",
- "family": "nvidia",
- "created_at": "2024-10-15 02:00:00 +0200",
- "context_window": 131072,
- "max_output_tokens": 16384,
+ "family": "openai",
+ "created_at": "2023-09-28 10:00:00 +1000",
+ "context_window": 4095,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -22262,20 +22772,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.12,
- "output_per_million": 0.3
+ "input_per_million": 1.5,
+ "output_per_million": 2.0
}
}
},
"metadata": {
- "description": "NVIDIA's Llama 3.1 Nemotron 70B is a language model designed for generating precise and useful responses. Leveraging [Llama 3.1 70B](/models/meta-llama/llama-3.1-70b-instruct) architecture and Reinforcement Learning from Human Feedback (RLHF), it excels in automatic alignment benchmarks. This model is tailored for applications requiring high accuracy in helpfulness and response generation, suitable for diverse user queries across multiple domains.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).",
+ "description": "This model is a variant of GPT-3.5 Turbo tuned for instructional prompts and omitting chat-related optimizations. Training data: up to Sep 2021.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -22284,41 +22792,39 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "GPT",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 16384,
- "is_moderated": false
+ "context_length": 4095,
+ "max_completion_tokens": 4096,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
- "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "nvidia/llama-3.1-nemotron-ultra-253b-v1",
- "name": "NVIDIA: Llama 3.1 Nemotron Ultra 253B v1",
+ "id": "openai/gpt-4",
+ "name": "OpenAI: GPT-4",
"provider": "openrouter",
- "family": "nvidia",
- "created_at": "2025-04-08 14:24:19 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "family": "openai",
+ "created_at": "2023-05-28 10:00:00 +1000",
+ "context_window": 8191,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -22330,18 +22836,19 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.6,
- "output_per_million": 1.7999999999999998
+ "input_per_million": 30.0,
+ "output_per_million": 60.0
}
}
},
"metadata": {
- "description": "Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) optimized for advanced reasoning, human-interactive chat, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta’s Llama-3.1-405B-Instruct, it has been significantly customized using Neural Architecture Search (NAS), resulting in enhanced efficiency, reduced memory usage, and improved inference latency. The model supports a context length of up to 128K tokens and can operate efficiently on an 8x NVIDIA H100 node.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.",
+ "description": "OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning capabilities. Training data: up to Sep 2021.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -22350,40 +22857,41 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
+ "tokenizer": "GPT",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
- "is_moderated": false
+ "context_length": 8191,
+ "max_completion_tokens": 4096,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"presence_penalty",
- "reasoning",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "nvidia/llama-3.1-nemotron-ultra-253b-v1:free",
- "name": "NVIDIA: Llama 3.1 Nemotron Ultra 253B v1 (free)",
+ "id": "openai/gpt-4-0314",
+ "name": "OpenAI: GPT-4 (older v0314)",
"provider": "openrouter",
- "family": "nvidia",
- "created_at": "2025-04-08 14:24:19 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "family": "openai",
+ "created_at": "2023-05-28 10:00:00 +1000",
+ "context_window": 8191,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -22395,11 +22903,19 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 30.0,
+ "output_per_million": 60.0
+ }
+ }
+ },
"metadata": {
- "description": "Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) optimized for advanced reasoning, human-interactive chat, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta’s Llama-3.1-405B-Instruct, it has been significantly customized using Neural Architecture Search (NAS), resulting in enhanced efficiency, reduced memory usage, and improved inference latency. The model supports a context length of up to 128K tokens and can operate efficiently on an 8x NVIDIA H100 node.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.",
+ "description": "GPT-4-0314 is the first version of GPT-4 released, with a context length of 8,192 tokens, and was supported until June 14. Training data: up to Sep 2021.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -22408,13 +22924,13 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
+ "tokenizer": "GPT",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
- "is_moderated": false
+ "context_length": 8191,
+ "max_completion_tokens": 4096,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
@@ -22422,26 +22938,27 @@
"logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "nvidia/llama-3.3-nemotron-super-49b-v1",
- "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1",
+ "id": "openai/gpt-4-1106-preview",
+ "name": "OpenAI: GPT-4 Turbo (older v1106)",
"provider": "openrouter",
- "family": "nvidia",
- "created_at": "2025-04-08 15:38:14 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "family": "openai",
+ "created_at": "2023-11-06 11:00:00 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -22453,18 +22970,19 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.13,
- "output_per_million": 0.39999999999999997
+ "input_per_million": 10.0,
+ "output_per_million": 30.0
}
}
},
"metadata": {
- "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.",
+ "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to April 2023.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -22473,13 +22991,13 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "GPT",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
- "is_moderated": false
+ "context_length": 128000,
+ "max_completion_tokens": 4096,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
@@ -22488,23 +23006,26 @@
"logprobs",
"max_tokens",
"presence_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "openai/chatgpt-4o-latest",
- "name": "OpenAI: ChatGPT-4o",
+ "id": "openai/gpt-4-turbo",
+ "name": "OpenAI: GPT-4 Turbo",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-08-14 02:00:00 +0200",
+ "created_at": "2024-04-09 10:00:00 +1000",
"context_window": 128000,
- "max_output_tokens": 16384,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -22517,18 +23038,19 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 5.0,
- "output_per_million": 15.0
+ "input_per_million": 10.0,
+ "output_per_million": 30.0
}
}
},
"metadata": {
- "description": "OpenAI ChatGPT 4o is continually updated by OpenAI to point to the current version of GPT-4o used by ChatGPT. It therefore differs slightly from the API version of [GPT-4o](/models/openai/gpt-4o) in that it has additional RLHF. It is intended for research and evaluation.\n\nOpenAI notes that this model is not suited for production use-cases as it may be removed or redirected to another model in the future.",
+ "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to December 2023.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -22543,7 +23065,7 @@
},
"top_provider": {
"context_length": 128000,
- "max_completion_tokens": 16384,
+ "max_completion_tokens": 4096,
"is_moderated": true
},
"per_request_limits": null,
@@ -22558,23 +23080,24 @@
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "openai/codex-mini",
- "name": "OpenAI: Codex Mini",
+ "id": "openai/gpt-4-turbo-preview",
+ "name": "OpenAI: GPT-4 Turbo Preview",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-05-16 17:36:01 +0200",
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "created_at": "2024-01-25 11:00:00 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -22589,18 +23112,16 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.5,
- "output_per_million": 6.0,
- "cached_input_per_million": 0.375
+ "input_per_million": 10.0,
+ "output_per_million": 30.0
}
}
},
"metadata": {
- "description": "codex-mini-latest is a fine-tuned version of o4-mini specifically for use in Codex CLI. For direct use in the API, we recommend starting with gpt-4.1.",
+ "description": "The preview GPT-4 model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Dec 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
@@ -22610,35 +23131,43 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 100000,
+ "context_length": 128000,
+ "max_completion_tokens": 4096,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
- "reasoning",
+ "presence_penalty",
"response_format",
"seed",
+ "stop",
"structured_outputs",
+ "temperature",
"tool_choice",
- "tools"
+ "tools",
+ "top_logprobs",
+ "top_p"
]
}
},
{
- "id": "openai/gpt-3.5-turbo",
- "name": "OpenAI: GPT-3.5 Turbo",
+ "id": "openai/gpt-4.1",
+ "name": "OpenAI: GPT-4.1",
"provider": "openrouter",
"family": "openai",
- "created_at": "2023-05-28 02:00:00 +0200",
- "context_window": 16385,
- "max_output_tokens": 4096,
+ "created_at": "2025-04-15 03:23:05 +1000",
+ "context_window": 1047576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output": [
"text"
@@ -22652,17 +23181,20 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.0,
+ "output_per_million": 8.0,
+ "cached_input_per_million": 0.5
}
}
},
"metadata": {
- "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.",
+ "description": "GPT-4.1 is a flagship large language model optimized for advanced instruction following, real-world software engineering, and long-context reasoning. It supports a 1 million token context window and outperforms GPT-4o and GPT-4.5 across coding (54.6% SWE-bench Verified), instruction compliance (87.4% IFEval), and multimodal understanding benchmarks. It is tuned for precise code diffs, agent reliability, and high recall in large document contexts, making it ideal for agents, IDE tooling, and enterprise knowledge retrieval.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output_modalities": [
"text"
@@ -22671,8 +23203,8 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 16385,
- "max_completion_tokens": 4096,
+ "context_length": 1047576,
+ "max_completion_tokens": 32768,
"is_moderated": true
},
"per_request_limits": null,
@@ -22690,22 +23222,25 @@
"tool_choice",
"tools",
"top_logprobs",
- "top_p"
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/gpt-3.5-turbo-0613",
- "name": "OpenAI: GPT-3.5 Turbo (older v0613)",
+ "id": "openai/gpt-4.1-mini",
+ "name": "OpenAI: GPT-4.1 Mini",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-01-25 01:00:00 +0100",
- "context_window": 4095,
- "max_output_tokens": 4096,
+ "created_at": "2025-04-15 03:23:01 +1000",
+ "context_window": 1047576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output": [
"text"
@@ -22719,17 +23254,20 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.0,
- "output_per_million": 2.0
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 1.5999999999999999,
+ "cached_input_per_million": 0.09999999999999999
}
}
},
"metadata": {
- "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.",
+ "description": "GPT-4.1 Mini is a mid-sized model delivering performance competitive with GPT-4o at substantially lower latency and cost. It retains a 1 million token context window and scores 45.1% on hard instruction evals, 35.8% on MultiChallenge, and 84.1% on IFEval. Mini also shows strong coding ability (e.g., 31.6% on Aider’s polyglot diff benchmark) and vision understanding, making it suitable for interactive applications with tight performance constraints.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output_modalities": [
"text"
@@ -22738,9 +23276,9 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 4095,
- "max_completion_tokens": 4096,
- "is_moderated": false
+ "context_length": 1047576,
+ "max_completion_tokens": 32768,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
@@ -22757,22 +23295,25 @@
"tool_choice",
"tools",
"top_logprobs",
- "top_p"
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/gpt-3.5-turbo-16k",
- "name": "OpenAI: GPT-3.5 Turbo 16k",
+ "id": "openai/gpt-4.1-nano",
+ "name": "OpenAI: GPT-4.1 Nano",
"provider": "openrouter",
"family": "openai",
- "created_at": "2023-08-28 02:00:00 +0200",
- "context_window": 16385,
- "max_output_tokens": 4096,
+ "created_at": "2025-04-15 03:22:49 +1000",
+ "context_window": 1047576,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output": [
"text"
@@ -22786,17 +23327,20 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 4.0
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.39999999999999997,
+ "cached_input_per_million": 0.024999999999999998
}
}
},
"metadata": {
- "description": "This model offers four times the context length of gpt-3.5-turbo, allowing it to support approximately 20 pages of text in a single request at a higher cost. Training data: up to Sep 2021.",
+ "description": "For tasks that demand low latency, GPT‑4.1 nano is the fastest and cheapest model in the GPT-4.1 series. It delivers exceptional performance at a small size with its 1 million token context window, and scores 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding – even higher than GPT‑4o mini. It’s ideal for tasks like classification or autocompletion.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "image",
+ "text",
+ "file"
],
"output_modalities": [
"text"
@@ -22805,8 +23349,8 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 16385,
- "max_completion_tokens": 4096,
+ "context_length": 1047576,
+ "max_completion_tokens": 32768,
"is_moderated": true
},
"per_request_limits": null,
@@ -22829,17 +23373,19 @@
}
},
{
- "id": "openai/gpt-3.5-turbo-instruct",
- "name": "OpenAI: GPT-3.5 Turbo Instruct",
+ "id": "openai/gpt-4o",
+ "name": "OpenAI: GPT-4o",
"provider": "openrouter",
"family": "openai",
- "created_at": "2023-09-28 02:00:00 +0200",
- "context_window": 4095,
- "max_output_tokens": 4096,
+ "created_at": "2024-05-13 10:00:00 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
@@ -22847,32 +23393,36 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.5,
- "output_per_million": 2.0
+ "input_per_million": 2.5,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 1.25
}
}
},
"metadata": {
- "description": "This model is a variant of GPT-3.5 Turbo tuned for instructional prompts and omitting chat-related optimizations. Training data: up to Sep 2021.",
+ "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
],
"tokenizer": "GPT",
- "instruct_type": "chatml"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 4095,
- "max_completion_tokens": 4096,
+ "context_length": 128000,
+ "max_completion_tokens": 16384,
"is_moderated": true
},
"per_request_limits": null,
@@ -22887,23 +23437,28 @@
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_logprobs",
- "top_p"
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/gpt-4",
- "name": "OpenAI: GPT-4",
+ "id": "openai/gpt-4o-2024-05-13",
+ "name": "OpenAI: GPT-4o (2024-05-13)",
"provider": "openrouter",
"family": "openai",
- "created_at": "2023-05-28 02:00:00 +0200",
- "context_window": 8191,
+ "created_at": "2024-05-13 10:00:00 +1000",
+ "context_window": 128000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
@@ -22917,17 +23472,19 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 30.0,
- "output_per_million": 60.0
+ "input_per_million": 5.0,
+ "output_per_million": 15.0
}
}
},
"metadata": {
- "description": "OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning capabilities. Training data: up to Sep 2021.",
+ "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
@@ -22936,7 +23493,7 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 8191,
+ "context_length": 128000,
"max_completion_tokens": 4096,
"is_moderated": true
},
@@ -22955,22 +23512,25 @@
"tool_choice",
"tools",
"top_logprobs",
- "top_p"
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/gpt-4-0314",
- "name": "OpenAI: GPT-4 (older v0314)",
+ "id": "openai/gpt-4o-2024-08-06",
+ "name": "OpenAI: GPT-4o (2024-08-06)",
"provider": "openrouter",
"family": "openai",
- "created_at": "2023-05-28 02:00:00 +0200",
- "context_window": 8191,
- "max_output_tokens": 4096,
+ "created_at": "2024-08-06 10:00:00 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
@@ -22984,17 +23544,20 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 30.0,
- "output_per_million": 60.0
+ "input_per_million": 2.5,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 1.25
}
}
},
"metadata": {
- "description": "GPT-4-0314 is the first version of GPT-4 released, with a context length of 8,192 tokens, and was supported until June 14. Training data: up to Sep 2021.",
+ "description": "The 2024-08-06 version of GPT-4o offers improved performance in structured outputs, with the ability to supply a JSON schema in the respone_format. Read more [here](https://openai.com/index/introducing-structured-outputs-in-the-api/).\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
@@ -23003,9 +23566,9 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 8191,
- "max_completion_tokens": 4096,
- "is_moderated": true
+ "context_length": 128000,
+ "max_completion_tokens": 16384,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
@@ -23022,22 +23585,25 @@
"tool_choice",
"tools",
"top_logprobs",
- "top_p"
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/gpt-4-1106-preview",
- "name": "OpenAI: GPT-4 Turbo (older v1106)",
+ "id": "openai/gpt-4o-2024-11-20",
+ "name": "OpenAI: GPT-4o (2024-11-20)",
"provider": "openrouter",
"family": "openai",
- "created_at": "2023-11-06 01:00:00 +0100",
+ "created_at": "2024-11-21 05:33:14 +1100",
"context_window": 128000,
- "max_output_tokens": 4096,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
@@ -23051,17 +23617,20 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 10.0,
- "output_per_million": 30.0
+ "input_per_million": 2.5,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 1.25
}
}
},
"metadata": {
- "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to April 2023.",
+ "description": "The 2024-11-20 version of GPT-4o offers a leveled-up creative writing ability with more natural, engaging, and tailored writing to improve relevance & readability. It’s also better at working with uploaded files, providing deeper insights & more thorough responses.\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
@@ -23071,7 +23640,7 @@
},
"top_provider": {
"context_length": 128000,
- "max_completion_tokens": 4096,
+ "max_completion_tokens": 16384,
"is_moderated": true
},
"per_request_limits": null,
@@ -23089,23 +23658,24 @@
"tool_choice",
"tools",
"top_logprobs",
- "top_p"
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/gpt-4-turbo",
- "name": "OpenAI: GPT-4 Turbo",
+ "id": "openai/gpt-4o-audio-preview",
+ "name": "OpenAI: GPT-4o Audio",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-04-09 02:00:00 +0200",
+ "created_at": "2025-08-15 14:44:21 +1000",
"context_window": 128000,
- "max_output_tokens": 4096,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "audio",
+ "text"
],
"output": [
"text"
@@ -23119,18 +23689,18 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 10.0,
- "output_per_million": 30.0
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to December 2023.",
+ "description": "The gpt-4o-audio-preview model adds support for audio inputs as prompts. This enhancement allows the model to detect nuances within audio recordings and add depth to generated user experiences. Audio outputs are currently not supported. Audio tokens are priced at $40 per million input audio tokens.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "audio",
+ "text"
],
"output_modalities": [
"text"
@@ -23140,7 +23710,7 @@
},
"top_provider": {
"context_length": 128000,
- "max_completion_tokens": 4096,
+ "max_completion_tokens": 16384,
"is_moderated": true
},
"per_request_limits": null,
@@ -23163,17 +23733,19 @@
}
},
{
- "id": "openai/gpt-4-turbo-preview",
- "name": "OpenAI: GPT-4 Turbo Preview",
+ "id": "openai/gpt-4o-mini",
+ "name": "OpenAI: GPT-4o-mini",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-01-25 01:00:00 +0100",
+ "created_at": "2024-07-18 10:00:00 +1000",
"context_window": 128000,
- "max_output_tokens": 4096,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
@@ -23187,17 +23759,20 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 10.0,
- "output_per_million": 30.0
+ "input_per_million": 0.15,
+ "output_per_million": 0.6,
+ "cached_input_per_million": 0.075
}
}
},
"metadata": {
- "description": "The preview GPT-4 model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Dec 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.",
+ "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.\n\n#multimodal",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
@@ -23207,7 +23782,7 @@
},
"top_provider": {
"context_length": 128000,
- "max_completion_tokens": 4096,
+ "max_completion_tokens": 16384,
"is_moderated": true
},
"per_request_limits": null,
@@ -23225,23 +23800,24 @@
"tool_choice",
"tools",
"top_logprobs",
- "top_p"
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/gpt-4.1",
- "name": "OpenAI: GPT-4.1",
+ "id": "openai/gpt-4o-mini-2024-07-18",
+ "name": "OpenAI: GPT-4o-mini (2024-07-18)",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-04-14 19:23:05 +0200",
- "context_window": 1047576,
- "max_output_tokens": 32768,
+ "created_at": "2024-07-18 10:00:00 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text",
+ "image",
"file"
],
"output": [
@@ -23254,21 +23830,21 @@
"structured_output"
],
"pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 2.0,
- "output_per_million": 8.0,
- "cached_input_per_million": 0.5
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.15,
+ "output_per_million": 0.6,
+ "cached_input_per_million": 0.075
}
}
},
"metadata": {
- "description": "GPT-4.1 is a flagship large language model optimized for advanced instruction following, real-world software engineering, and long-context reasoning. It supports a 1 million token context window and outperforms GPT-4o and GPT-4.5 across coding (54.6% SWE-bench Verified), instruction compliance (87.4% IFEval), and multimodal understanding benchmarks. It is tuned for precise code diffs, agent reliability, and high recall in large document contexts, making it ideal for agents, IDE tooling, and enterprise knowledge retrieval.",
+ "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.\n\n#multimodal",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "image",
"text",
+ "image",
"file"
],
"output_modalities": [
@@ -23278,8 +23854,8 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 1047576,
- "max_completion_tokens": 32768,
+ "context_length": 128000,
+ "max_completion_tokens": 16384,
"is_moderated": true
},
"per_request_limits": null,
@@ -23303,19 +23879,17 @@
}
},
{
- "id": "openai/gpt-4.1-mini",
- "name": "OpenAI: GPT-4.1 Mini",
+ "id": "openai/gpt-4o-mini-search-preview",
+ "name": "OpenAI: GPT-4o-mini Search Preview",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-04-14 19:23:01 +0200",
- "context_window": 1047576,
- "max_output_tokens": 32768,
+ "created_at": "2025-03-13 09:22:02 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
- "text",
- "file"
+ "text"
],
"output": [
"text"
@@ -23323,26 +23897,22 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 1.5999999999999999,
- "cached_input_per_million": 0.09999999999999999
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
}
}
},
"metadata": {
- "description": "GPT-4.1 Mini is a mid-sized model delivering performance competitive with GPT-4o at substantially lower latency and cost. It retains a 1 million token context window and scores 45.1% on hard instruction evals, 35.8% on MultiChallenge, and 84.1% on IFEval. Mini also shows strong coding ability (e.g., 31.6% on Aider’s polyglot diff benchmark) and vision understanding, making it suitable for interactive applications with tight performance constraints.",
+ "description": "GPT-4o mini Search Preview is a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
- "text",
- "file"
+ "text"
],
"output_modalities": [
"text"
@@ -23351,44 +23921,31 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 1047576,
- "max_completion_tokens": 32768,
+ "context_length": 128000,
+ "max_completion_tokens": 16384,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "presence_penalty",
"response_format",
- "seed",
- "stop",
"structured_outputs",
- "temperature",
- "tool_choice",
- "tools",
- "top_logprobs",
- "top_p",
"web_search_options"
]
}
},
{
- "id": "openai/gpt-4.1-nano",
- "name": "OpenAI: GPT-4.1 Nano",
+ "id": "openai/gpt-4o-search-preview",
+ "name": "OpenAI: GPT-4o Search Preview",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-04-14 19:22:49 +0200",
- "context_window": 1047576,
- "max_output_tokens": 32768,
+ "created_at": "2025-03-13 09:19:09 +1100",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
- "text",
- "file"
+ "text"
],
"output": [
"text"
@@ -23396,26 +23953,22 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.39999999999999997,
- "cached_input_per_million": 0.024999999999999998
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "For tasks that demand low latency, GPT‑4.1 nano is the fastest and cheapest model in the GPT-4.1 series. It delivers exceptional performance at a small size with its 1 million token context window, and scores 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding – even higher than GPT‑4o mini. It’s ideal for tasks like classification or autocompletion.",
+ "description": "GPT-4o Search Previewis a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
- "text",
- "file"
+ "text"
],
"output_modalities": [
"text"
@@ -23424,37 +23977,27 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 1047576,
- "max_completion_tokens": 32768,
+ "context_length": 128000,
+ "max_completion_tokens": 16384,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "presence_penalty",
"response_format",
- "seed",
- "stop",
"structured_outputs",
- "temperature",
- "tool_choice",
- "tools",
- "top_logprobs",
- "top_p"
+ "web_search_options"
]
}
},
{
- "id": "openai/gpt-4o",
- "name": "OpenAI: GPT-4o",
+ "id": "openai/gpt-4o:extended",
+ "name": "OpenAI: GPT-4o (extended)",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-05-13 02:00:00 +0200",
+ "created_at": "2024-05-13 10:00:00 +1000",
"context_window": 128000,
- "max_output_tokens": 16384,
+ "max_output_tokens": 64000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -23474,9 +24017,8 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0,
- "cached_input_per_million": 1.25
+ "input_per_million": 6.0,
+ "output_per_million": 18.0
}
}
},
@@ -23497,7 +24039,7 @@
},
"top_provider": {
"context_length": 128000,
- "max_completion_tokens": 16384,
+ "max_completion_tokens": 64000,
"is_moderated": true
},
"per_request_limits": null,
@@ -23521,13 +24063,13 @@
}
},
{
- "id": "openai/gpt-4o-2024-05-13",
- "name": "OpenAI: GPT-4o (2024-05-13)",
+ "id": "openai/gpt-5",
+ "name": "OpenAI: GPT-5",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-05-13 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 4096,
+ "created_at": "2025-08-08 03:23:33 +1000",
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -23547,13 +24089,14 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 5.0,
- "output_per_million": 15.0
+ "input_per_million": 1.25,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 0.125
}
}
},
"metadata": {
- "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal",
+ "description": "GPT-5 is OpenAI’s most advanced model, offering major improvements in reasoning, code quality, and user experience. It is optimized for complex tasks that require step-by-step reasoning, instruction following, and accuracy in high-stakes use cases. It supports test-time routing features and advanced prompt understanding, including user-specified intent like \"think hard about this.\" Improvements include reductions in hallucination, sycophancy, and better performance in coding, writing, and health-related tasks.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -23568,44 +24111,37 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4096,
+ "context_length": 400000,
+ "max_completion_tokens": 128000,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
+ "include_reasoning",
"max_tokens",
- "presence_penalty",
+ "reasoning",
"response_format",
"seed",
- "stop",
"structured_outputs",
- "temperature",
"tool_choice",
- "tools",
- "top_logprobs",
- "top_p",
- "web_search_options"
+ "tools"
]
}
},
{
- "id": "openai/gpt-4o-2024-08-06",
- "name": "OpenAI: GPT-4o (2024-08-06)",
+ "id": "openai/gpt-5-chat",
+ "name": "OpenAI: GPT-5 Chat",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-08-06 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "created_at": "2025-08-08 03:30:37 +1000",
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
+ "file",
"image",
- "file"
+ "text"
],
"output": [
"text"
@@ -23613,26 +24149,25 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
+ "input_per_million": 1.25,
"output_per_million": 10.0,
- "cached_input_per_million": 1.25
+ "cached_input_per_million": 0.125
}
}
},
"metadata": {
- "description": "The 2024-08-06 version of GPT-4o offers improved performance in structured outputs, with the ability to supply a JSON schema in the respone_format. Read more [here](https://openai.com/index/introducing-structured-outputs-in-the-api/).\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)",
+ "description": "GPT-5 Chat is designed for advanced, natural, multimodal, and context-aware conversations for enterprise applications.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "text",
+ "file",
"image",
- "file"
+ "text"
],
"output_modalities": [
"text"
@@ -23641,38 +24176,29 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 16384,
- "is_moderated": false
+ "context_length": 400000,
+ "max_completion_tokens": 128000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
+ "include_reasoning",
"max_tokens",
- "presence_penalty",
+ "reasoning",
"response_format",
"seed",
- "stop",
- "structured_outputs",
- "temperature",
- "tool_choice",
- "tools",
- "top_logprobs",
- "top_p",
- "web_search_options"
+ "structured_outputs"
]
}
},
{
- "id": "openai/gpt-4o-2024-11-20",
- "name": "OpenAI: GPT-4o (2024-11-20)",
+ "id": "openai/gpt-5-mini",
+ "name": "OpenAI: GPT-5 Mini",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-11-20 19:33:14 +0100",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "created_at": "2025-08-08 03:23:27 +1000",
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -23692,14 +24218,14 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0,
- "cached_input_per_million": 1.25
+ "input_per_million": 0.25,
+ "output_per_million": 2.0,
+ "cached_input_per_million": 0.024999999999999998
}
}
},
"metadata": {
- "description": "The 2024-11-20 version of GPT-4o offers a leveled-up creative writing ability with more natural, engaging, and tailored writing to improve relevance & readability. It’s also better at working with uploaded files, providing deeper insights & more thorough responses.\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.",
+ "description": "GPT-5 Mini is a compact version of GPT-5, designed to handle lighter-weight reasoning tasks. It provides the same instruction-following and safety-tuning benefits as GPT-5, but with reduced latency and cost. GPT-5 Mini is the successor to OpenAI's o4-mini model.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -23714,43 +24240,37 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 16384,
+ "context_length": 400000,
+ "max_completion_tokens": 128000,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
+ "include_reasoning",
"max_tokens",
- "presence_penalty",
+ "reasoning",
"response_format",
"seed",
- "stop",
"structured_outputs",
- "temperature",
"tool_choice",
- "tools",
- "top_logprobs",
- "top_p",
- "web_search_options"
+ "tools"
]
}
},
{
- "id": "openai/gpt-4o-audio-preview",
- "name": "OpenAI: GPT-4o Audio",
+ "id": "openai/gpt-5-nano",
+ "name": "OpenAI: GPT-5 Nano",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-08-15 06:44:21 +0200",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "created_at": "2025-08-08 03:23:22 +1000",
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
- "text"
+ "text",
+ "image",
+ "file"
],
"output": [
"text"
@@ -23764,18 +24284,20 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 0.049999999999999996,
+ "output_per_million": 0.39999999999999997,
+ "cached_input_per_million": 0.005
}
}
},
"metadata": {
- "description": "The gpt-4o-audio-preview model adds support for audio inputs as prompts. This enhancement allows the model to detect nuances within audio recordings and add depth to generated user experiences. Audio outputs are currently not supported. Audio tokens are priced at $40 per million input audio tokens.",
+ "description": "GPT-5-Nano is the smallest and fastest variant in the GPT-5 system, optimized for developer tools, rapid interactions, and ultra-low latency environments. While limited in reasoning depth compared to its larger counterparts, it retains key instruction-following and safety features. It is the successor to GPT-4.1-nano and offers a lightweight option for cost-sensitive or real-time applications.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "audio",
- "text"
+ "text",
+ "image",
+ "file"
],
"output_modalities": [
"text"
@@ -23784,43 +24306,35 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 16384,
+ "context_length": 400000,
+ "max_completion_tokens": 128000,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
+ "include_reasoning",
"max_tokens",
- "presence_penalty",
+ "reasoning",
"response_format",
"seed",
- "stop",
"structured_outputs",
- "temperature",
"tool_choice",
- "tools",
- "top_logprobs",
- "top_p"
+ "tools"
]
}
},
{
- "id": "openai/gpt-4o-mini",
- "name": "OpenAI: GPT-4o-mini",
+ "id": "openai/gpt-oss-120b",
+ "name": "OpenAI: gpt-oss-120b",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-07-18 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "created_at": "2025-08-06 03:17:11 +1000",
+ "context_window": 131000,
+ "max_output_tokens": 131000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
@@ -23829,25 +24343,23 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6,
- "cached_input_per_million": 0.075
+ "input_per_million": 0.072,
+ "output_per_million": 0.28
}
}
},
"metadata": {
- "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.\n\n#multimodal",
+ "description": "gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized to run on a single H100 GPU with native MXFP4 quantization. The model supports configurable reasoning depth, full chain-of-thought access, and native tool use, including function calling, browsing, and structured output generation.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
@@ -23856,17 +24368,21 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 16384,
- "is_moderated": true
+ "context_length": 131000,
+ "max_completion_tokens": 131000,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -23874,26 +24390,24 @@
"temperature",
"tool_choice",
"tools",
+ "top_k",
"top_logprobs",
- "top_p",
- "web_search_options"
+ "top_p"
]
}
},
{
- "id": "openai/gpt-4o-mini-2024-07-18",
- "name": "OpenAI: GPT-4o-mini (2024-07-18)",
+ "id": "openai/gpt-oss-20b",
+ "name": "OpenAI: gpt-oss-20b",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-07-18 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "created_at": "2025-08-06 03:17:09 +1000",
+ "context_window": 131000,
+ "max_output_tokens": 131000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
@@ -23902,25 +24416,23 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6,
- "cached_input_per_million": 0.075
+ "input_per_million": 0.04,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.\n\n#multimodal",
+ "description": "gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for lower-latency inference and deployability on consumer or single-GPU hardware. The model is trained in OpenAI’s Harmony response format and supports reasoning level configuration, fine-tuning, and agentic capabilities including function calling, tool use, and structured outputs.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
@@ -23929,17 +24441,21 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 16384,
- "is_moderated": true
+ "context_length": 131000,
+ "max_completion_tokens": 131000,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
@@ -23947,20 +24463,20 @@
"temperature",
"tool_choice",
"tools",
+ "top_k",
"top_logprobs",
- "top_p",
- "web_search_options"
+ "top_p"
]
}
},
{
- "id": "openai/gpt-4o-mini-search-preview",
- "name": "OpenAI: GPT-4o-mini Search Preview",
+ "id": "openai/gpt-oss-20b:free",
+ "name": "OpenAI: gpt-oss-20b (free)",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-03-12 23:22:02 +0100",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "created_at": "2025-08-06 03:17:09 +1000",
+ "context_window": 131072,
+ "max_output_tokens": 131072,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -23974,16 +24490,9 @@
"streaming",
"structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "GPT-4o mini Search Preview is a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.",
+ "description": "gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for lower-latency inference and deployability on consumer or single-GPU hardware. The model is trained in OpenAI’s Harmony response format and supports reasoning level configuration, fine-tuning, and agentic capabilities including function calling, tool use, and structured outputs.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -23996,31 +24505,35 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 16384,
- "is_moderated": true
+ "context_length": 131072,
+ "max_completion_tokens": 131072,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "include_reasoning",
"max_tokens",
+ "reasoning",
"response_format",
"structured_outputs",
- "web_search_options"
+ "temperature",
+ "top_p"
]
}
},
{
- "id": "openai/gpt-4o-search-preview",
- "name": "OpenAI: GPT-4o Search Preview",
+ "id": "openai/o1",
+ "name": "OpenAI: o1",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-03-12 23:19:09 +0100",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "created_at": "2024-12-18 05:26:39 +1100",
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -24028,22 +24541,25 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 15.0,
+ "output_per_million": 60.0,
+ "cached_input_per_million": 7.5
}
}
},
"metadata": {
- "description": "GPT-4o Search Previewis a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.",
+ "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding. The o1 model series is trained with large-scale reinforcement learning to reason using chain of thought. \n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -24052,59 +24568,56 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 16384,
+ "context_length": 200000,
+ "max_completion_tokens": 100000,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
"max_tokens",
"response_format",
+ "seed",
"structured_outputs",
- "web_search_options"
+ "tool_choice",
+ "tools"
]
}
},
{
- "id": "openai/gpt-4o:extended",
- "name": "OpenAI: GPT-4o (extended)",
+ "id": "openai/o1-mini",
+ "name": "OpenAI: o1-mini",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-05-13 02:00:00 +0200",
+ "created_at": "2024-09-12 10:00:00 +1000",
"context_window": 128000,
- "max_output_tokens": 64000,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 6.0,
- "output_per_million": 18.0
+ "input_per_million": 1.1,
+ "output_per_million": 4.4,
+ "cached_input_per_million": 0.55
}
}
},
"metadata": {
- "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal",
+ "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
@@ -24114,70 +24627,51 @@
},
"top_provider": {
"context_length": 128000,
- "max_completion_tokens": 64000,
+ "max_completion_tokens": 65536,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "presence_penalty",
- "response_format",
- "seed",
- "stop",
- "structured_outputs",
- "temperature",
- "tool_choice",
- "tools",
- "top_logprobs",
- "top_p",
- "web_search_options"
+ "seed"
]
}
},
{
- "id": "openai/gpt-5",
- "name": "OpenAI: GPT-5",
+ "id": "openai/o1-mini-2024-09-12",
+ "name": "OpenAI: o1-mini (2024-09-12)",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-08-07 19:23:33 +0200",
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "created_at": "2024-09-12 10:00:00 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.125
+ "input_per_million": 1.1,
+ "output_per_million": 4.4,
+ "cached_input_per_million": 0.55
}
}
},
"metadata": {
- "description": "GPT-5 is OpenAI’s most advanced model, offering major improvements in reasoning, code quality, and user experience. It is optimized for complex tasks that require step-by-step reasoning, instruction following, and accuracy in high-stakes use cases. It supports test-time routing features and advanced prompt understanding, including user-specified intent like \"think hard about this.\" Improvements include reductions in hallucination, sycophancy, and better performance in coding, writing, and health-related tasks.",
+ "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
@@ -24186,37 +24680,30 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 400000,
- "max_completion_tokens": 128000,
+ "context_length": 128000,
+ "max_completion_tokens": 65536,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
"max_tokens",
- "reasoning",
- "response_format",
- "seed",
- "structured_outputs",
- "tool_choice",
- "tools"
+ "seed"
]
}
},
{
- "id": "openai/gpt-5-chat",
- "name": "OpenAI: GPT-5 Chat",
+ "id": "openai/o1-pro",
+ "name": "OpenAI: o1-pro",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-08-07 19:30:37 +0200",
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "created_at": "2025-03-20 09:26:51 +1100",
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "file",
- "image",
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -24229,20 +24716,18 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.125
+ "input_per_million": 150.0,
+ "output_per_million": 600.0
}
}
},
"metadata": {
- "description": "GPT-5 Chat is designed for advanced, natural, multimodal, and context-aware conversations for enterprise applications.",
+ "description": "The o1 series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o1-pro model uses more compute to think harder and provide consistently better answers.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "file",
- "image",
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -24251,8 +24736,8 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 400000,
- "max_completion_tokens": 128000,
+ "context_length": 200000,
+ "max_completion_tokens": 100000,
"is_moderated": true
},
"per_request_limits": null,
@@ -24267,18 +24752,18 @@
}
},
{
- "id": "openai/gpt-5-mini",
- "name": "OpenAI: GPT-5 Mini",
+ "id": "openai/o3",
+ "name": "OpenAI: o3",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-08-07 19:23:27 +0200",
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "created_at": "2025-04-17 03:10:57 +1000",
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
+ "text",
"file"
],
"output": [
@@ -24293,19 +24778,19 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.25,
- "output_per_million": 2.0,
- "cached_input_per_million": 0.024999999999999998
+ "input_per_million": 2.0,
+ "output_per_million": 8.0,
+ "cached_input_per_million": 0.5
}
}
},
"metadata": {
- "description": "GPT-5 Mini is a compact version of GPT-5, designed to handle lighter-weight reasoning tasks. It provides the same instruction-following and safety-tuning benefits as GPT-5, but with reduced latency and cost. GPT-5 Mini is the successor to OpenAI's o4-mini model.",
+ "description": "o3 is a well-rounded and powerful model across domains. It sets a new standard for math, science, coding, and visual reasoning tasks. It also excels at technical writing and instruction-following. Use it to think through multi-step problems that involve analysis across text, code, and images. Note that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "text",
"image",
+ "text",
"file"
],
"output_modalities": [
@@ -24315,8 +24800,8 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 400000,
- "max_completion_tokens": 128000,
+ "context_length": 200000,
+ "max_completion_tokens": 100000,
"is_moderated": true
},
"per_request_limits": null,
@@ -24333,19 +24818,17 @@
}
},
{
- "id": "openai/gpt-5-nano",
- "name": "OpenAI: GPT-5 Nano",
+ "id": "openai/o3-mini",
+ "name": "OpenAI: o3 Mini",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-08-07 19:23:22 +0200",
- "context_window": 400000,
- "max_output_tokens": 128000,
+ "created_at": "2025-02-01 06:28:41 +1100",
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file"
+ "text"
],
"output": [
"text"
@@ -24359,37 +24842,33 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.049999999999999996,
- "output_per_million": 0.39999999999999997,
- "cached_input_per_million": 0.005
+ "input_per_million": 1.1,
+ "output_per_million": 4.4,
+ "cached_input_per_million": 0.55
}
}
},
"metadata": {
- "description": "GPT-5-Nano is the smallest and fastest variant in the GPT-5 system, optimized for developer tools, rapid interactions, and ultra-low latency environments. While limited in reasoning depth compared to its larger counterparts, it retains key instruction-following and safety features. It is the successor to GPT-4.1-nano and offers a lightweight option for cost-sensitive or real-time applications.",
+ "description": "OpenAI o3-mini is a cost-efficient language model optimized for STEM reasoning tasks, particularly excelling in science, mathematics, and coding.\n\nThis model supports the `reasoning_effort` parameter, which can be set to \"high\", \"medium\", or \"low\" to control the thinking time of the model. The default is \"medium\". OpenRouter also offers the model slug `openai/o3-mini-high` to default the parameter to \"high\".\n\nThe model features three adjustable reasoning effort levels and supports key developer capabilities including function calling, structured outputs, and streaming, though it does not include vision processing capabilities.\n\nThe model demonstrates significant improvements over its predecessor, with expert testers preferring its responses 56% of the time and noting a 39% reduction in major errors on complex questions. With medium reasoning effort settings, o3-mini matches the performance of the larger o1 model on challenging reasoning evaluations like AIME and GPQA, while maintaining lower latency and cost.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "GPT",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 400000,
- "max_completion_tokens": 128000,
+ "context_length": 200000,
+ "max_completion_tokens": 100000,
"is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
"max_tokens",
- "reasoning",
"response_format",
"seed",
"structured_outputs",
@@ -24399,13 +24878,13 @@
}
},
{
- "id": "openai/gpt-oss-120b",
- "name": "OpenAI: gpt-oss-120b",
+ "id": "openai/o3-mini-high",
+ "name": "OpenAI: o3 Mini High",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-08-05 19:17:11 +0200",
- "context_window": 131000,
- "max_output_tokens": 131000,
+ "created_at": "2025-02-13 02:03:31 +1100",
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -24418,19 +24897,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.072,
- "output_per_million": 0.28
+ "input_per_million": 1.1,
+ "output_per_million": 4.4,
+ "cached_input_per_million": 0.55
}
}
},
"metadata": {
- "description": "gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized to run on a single H100 GPU with native MXFP4 quantization. The model supports configurable reasoning depth, full chain-of-thought access, and native tool use, including function calling, browsing, and structured output generation.",
+ "description": "OpenAI o3-mini-high is the same model as [o3-mini](/openai/o3-mini) with reasoning_effort set to high. \n\no3-mini is a cost-efficient language model optimized for STEM reasoning tasks, particularly excelling in science, mathematics, and coding. The model features three adjustable reasoning effort levels and supports key developer capabilities including function calling, structured outputs, and streaming, though it does not include vision processing capabilities.\n\nThe model demonstrates significant improvements over its predecessor, with expert testers preferring its responses 56% of the time and noting a 39% reduction in major errors on complex questions. With medium reasoning effort settings, o3-mini matches the performance of the larger o1 model on challenging reasoning evaluations like AIME and GPQA, while maintaining lower latency and cost.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -24439,50 +24918,39 @@
"output_modalities": [
"text"
],
- "tokenizer": "GPT",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 131000,
- "max_completion_tokens": 131000,
- "is_moderated": false
+ "context_length": 200000,
+ "max_completion_tokens": 100000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
"response_format",
"seed",
- "stop",
"structured_outputs",
- "temperature",
"tool_choice",
- "tools",
- "top_k",
- "top_logprobs",
- "top_p"
+ "tools"
]
}
},
{
- "id": "openai/gpt-oss-20b",
- "name": "OpenAI: gpt-oss-20b",
+ "id": "openai/o3-pro",
+ "name": "OpenAI: o3 Pro",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-08-05 19:17:09 +0200",
- "context_window": 131000,
- "max_output_tokens": 131000,
+ "created_at": "2025-06-11 09:32:32 +1000",
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "file",
+ "image"
],
"output": [
"text"
@@ -24491,23 +24959,24 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.04,
- "output_per_million": 0.15
+ "input_per_million": 20.0,
+ "output_per_million": 80.0
}
}
},
"metadata": {
- "description": "gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for lower-latency inference and deployability on consumer or single-GPU hardware. The model is trained in OpenAI’s Harmony response format and supports reasoning level configuration, fine-tuning, and agentic capabilities including function calling, tool use, and structured outputs.",
+ "description": "The o-series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o3-pro model uses more compute to think harder and provide consistently better answers.\n\nNote that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "file",
+ "image"
],
"output_modalities": [
"text"
@@ -24516,45 +24985,35 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 131000,
- "max_completion_tokens": 131000,
- "is_moderated": false
+ "context_length": 200000,
+ "max_completion_tokens": 100000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
"response_format",
"seed",
- "stop",
"structured_outputs",
- "temperature",
"tool_choice",
- "tools",
- "top_k",
- "top_logprobs",
- "top_p"
+ "tools"
]
}
},
{
- "id": "openai/gpt-oss-20b:free",
- "name": "OpenAI: gpt-oss-20b (free)",
+ "id": "openai/o4-mini",
+ "name": "OpenAI: o4 Mini",
"provider": "openrouter",
"family": "openai",
- "created_at": "2025-08-05 19:17:09 +0200",
- "context_window": 131072,
- "max_output_tokens": 131072,
+ "created_at": "2025-04-17 02:29:02 +1000",
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -24563,26 +25022,36 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.1,
+ "output_per_million": 4.4,
+ "cached_input_per_million": 0.275
+ }
+ }
+ },
"metadata": {
- "description": "gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for lower-latency inference and deployability on consumer or single-GPU hardware. The model is trained in OpenAI’s Harmony response format and supports reasoning level configuration, fine-tuning, and agentic capabilities including function calling, tool use, and structured outputs.",
+ "description": "OpenAI o4-mini is a compact reasoning model in the o-series, optimized for fast, cost-efficient performance while retaining strong multimodal and agentic capabilities. It supports tool use and demonstrates competitive reasoning and coding performance across benchmarks like AIME (99.5% with Python) and SWE-bench, outperforming its predecessor o3-mini and even approaching o3 in some domains.\n\nDespite its smaller size, o4-mini exhibits high accuracy in STEM tasks, visual problem solving (e.g., MathVista, MMMU), and code editing. It is especially well-suited for high-throughput scenarios where latency or cost is critical. Thanks to its efficient architecture and refined reinforcement learning training, o4-mini can chain tools, generate structured outputs, and solve multi-step tasks with minimal delay—often in under a minute.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
+ "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "GPT",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 131072,
- "is_moderated": false
+ "context_length": 200000,
+ "max_completion_tokens": 100000,
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
@@ -24590,25 +25059,27 @@
"max_tokens",
"reasoning",
"response_format",
+ "seed",
"structured_outputs",
- "temperature",
- "top_p"
+ "tool_choice",
+ "tools"
]
}
},
{
- "id": "openai/o1",
- "name": "OpenAI: o1",
+ "id": "openai/o4-mini-high",
+ "name": "OpenAI: o4 Mini High",
"provider": "openrouter",
"family": "openai",
- "created_at": "2024-12-17 19:26:39 +0100",
+ "created_at": "2025-04-17 03:23:32 +1000",
"context_window": 200000,
"max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text",
- "image"
+ "file"
],
"output": [
"text"
@@ -24622,24 +25093,25 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 60.0,
- "cached_input_per_million": 7.5
+ "input_per_million": 1.1,
+ "output_per_million": 4.4,
+ "cached_input_per_million": 0.275
}
}
},
"metadata": {
- "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding. The o1 model series is trained with large-scale reinforcement learning to reason using chain of thought. \n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n",
+ "description": "OpenAI o4-mini-high is the same model as [o4-mini](/openai/o4-mini) with reasoning_effort set to high. \n\nOpenAI o4-mini is a compact reasoning model in the o-series, optimized for fast, cost-efficient performance while retaining strong multimodal and agentic capabilities. It supports tool use and demonstrates competitive reasoning and coding performance across benchmarks like AIME (99.5% with Python) and SWE-bench, outperforming its predecessor o3-mini and even approaching o3 in some domains.\n\nDespite its smaller size, o4-mini exhibits high accuracy in STEM tasks, visual problem solving (e.g., MathVista, MMMU), and code editing. It is especially well-suited for high-throughput scenarios where latency or cost is critical. Thanks to its efficient architecture and refined reinforcement learning training, o4-mini can chain tools, generate structured outputs, and solve multi-step tasks with minimal delay—often in under a minute.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
+ "image",
"text",
- "image"
+ "file"
],
"output_modalities": [
"text"
],
- "tokenizer": "GPT",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
@@ -24649,7 +25121,9 @@
},
"per_request_limits": null,
"supported_parameters": [
+ "include_reasoning",
"max_tokens",
+ "reasoning",
"response_format",
"seed",
"structured_outputs",
@@ -24659,16 +25133,17 @@
}
},
{
- "id": "openai/o1-mini",
- "name": "OpenAI: o1-mini",
+ "id": "opengvlab/internvl3-14b",
+ "name": "OpenGVLab: InternVL3 14B",
"provider": "openrouter",
- "family": "openai",
- "created_at": "2024-09-12 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 65536,
+ "family": "opengvlab",
+ "created_at": "2025-04-30 23:55:55 +1000",
+ "context_window": 12288,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -24681,45 +25156,46 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.1,
- "output_per_million": 4.4,
- "cached_input_per_million": 0.55
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.39999999999999997
}
}
},
"metadata": {
- "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.",
+ "description": "The 14b version of the InternVL3 series. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
+ "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "GPT",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 65536,
- "is_moderated": true
+ "context_length": 12288,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"max_tokens",
- "seed"
+ "temperature",
+ "top_p"
]
}
},
{
- "id": "openai/o1-mini-2024-09-12",
- "name": "OpenAI: o1-mini (2024-09-12)",
+ "id": "openrouter/auto",
+ "name": "Auto Router",
"provider": "openrouter",
- "family": "openai",
- "created_at": "2024-09-12 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 65536,
+ "family": "openrouter",
+ "created_at": "2023-11-08 11:00:00 +1100",
+ "context_window": 2000000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -24732,17 +25208,9 @@
"capabilities": [
"streaming"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 1.1,
- "output_per_million": 4.4,
- "cached_input_per_million": 0.55
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.",
+ "description": "Your prompt will be processed by a meta-model and routed to one of dozens of models (see below), optimizing for the best possible output.\n\nTo see which model was used, visit [Activity](/activity), or read the `model` attribute of the response. Your response will be priced at the same rate as the routed model.\n\nThe meta-model is powered by [Not Diamond](https://docs.notdiamond.ai/docs/how-not-diamond-works). Learn more in our [docs](/docs/model-routing).\n\nRequests will be routed to the following models:\n- [openai/gpt-4o-2024-08-06](/openai/gpt-4o-2024-08-06)\n- [openai/gpt-4o-2024-05-13](/openai/gpt-4o-2024-05-13)\n- [openai/gpt-4o-mini-2024-07-18](/openai/gpt-4o-mini-2024-07-18)\n- [openai/chatgpt-4o-latest](/openai/chatgpt-4o-latest)\n- [openai/o1-preview-2024-09-12](/openai/o1-preview-2024-09-12)\n- [openai/o1-mini-2024-09-12](/openai/o1-mini-2024-09-12)\n- [anthropic/claude-3.5-sonnet](/anthropic/claude-3.5-sonnet)\n- [anthropic/claude-3.5-haiku](/anthropic/claude-3.5-haiku)\n- [anthropic/claude-3-opus](/anthropic/claude-3-opus)\n- [anthropic/claude-2.1](/anthropic/claude-2.1)\n- [google/gemini-pro-1.5](/google/gemini-pro-1.5)\n- [google/gemini-flash-1.5](/google/gemini-flash-1.5)\n- [mistralai/mistral-large-2407](/mistralai/mistral-large-2407)\n- [mistralai/mistral-nemo](/mistralai/mistral-nemo)\n- [deepseek/deepseek-r1](/deepseek/deepseek-r1)\n- [meta-llama/llama-3.1-70b-instruct](/meta-llama/llama-3.1-70b-instruct)\n- [meta-llama/llama-3.1-405b-instruct](/meta-llama/llama-3.1-405b-instruct)\n- [mistralai/mixtral-8x22b-instruct](/mistralai/mixtral-8x22b-instruct)\n- [cohere/command-r-plus](/cohere/command-r-plus)\n- [cohere/command-r](/cohere/command-r)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -24751,34 +25219,30 @@
"output_modalities": [
"text"
],
- "tokenizer": "GPT",
+ "tokenizer": "Router",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 65536,
- "is_moderated": true
+ "context_length": null,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
- "supported_parameters": [
- "max_tokens",
- "seed"
- ]
+ "supported_parameters": []
}
},
{
- "id": "openai/o1-pro",
- "name": "OpenAI: o1-pro",
+ "id": "perplexity/r1-1776",
+ "name": "Perplexity: R1 1776",
"provider": "openrouter",
- "family": "openai",
- "created_at": "2025-03-19 23:26:51 +0100",
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "perplexity",
+ "created_at": "2025-02-20 09:42:09 +1100",
+ "context_window": 128000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -24786,120 +25250,119 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 150.0,
- "output_per_million": 600.0
+ "input_per_million": 2.0,
+ "output_per_million": 8.0
}
}
},
"metadata": {
- "description": "The o1 series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o1-pro model uses more compute to think harder and provide consistently better answers.",
+ "description": "R1 1776 is a version of DeepSeek-R1 that has been post-trained to remove censorship constraints related to topics restricted by the Chinese government. The model retains its original reasoning capabilities while providing direct responses to a wider range of queries. R1 1776 is an offline chat model that does not use the perplexity search subsystem.\n\nThe model was tested on a multilingual dataset of over 1,000 examples covering sensitive topics to measure its likelihood of refusal or overly filtered responses. [Evaluation Results](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/GiN2VqC5hawUgAGJ6oHla.png) Its performance on math and reasoning benchmarks remains similar to the base R1 model. [Reasoning Performance](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/n4Z9Byqp2S7sKUvCvI40R.png)\n\nRead more on the [Blog Post](https://perplexity.ai/hub/blog/open-sourcing-r1-1776)",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "GPT",
- "instruct_type": null
+ "tokenizer": "DeepSeek",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 100000,
- "is_moderated": true
+ "context_length": 128000,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"include_reasoning",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
"reasoning",
- "response_format",
- "seed",
- "structured_outputs"
+ "repetition_penalty",
+ "stop",
+ "temperature",
+ "top_k",
+ "top_p"
]
}
},
{
- "id": "openai/o3",
- "name": "OpenAI: o3",
+ "id": "perplexity/sonar",
+ "name": "Perplexity: Sonar",
"provider": "openrouter",
- "family": "openai",
- "created_at": "2025-04-16 19:10:57 +0200",
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "perplexity",
+ "created_at": "2025-01-28 08:36:48 +1100",
+ "context_window": 127072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text",
- "file"
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 8.0,
- "cached_input_per_million": 0.5
+ "input_per_million": 1.0,
+ "output_per_million": 1.0
}
}
},
"metadata": {
- "description": "o3 is a well-rounded and powerful model across domains. It sets a new standard for math, science, coding, and visual reasoning tasks. It also excels at technical writing and instruction-following. Use it to think through multi-step problems that involve analysis across text, code, and images. Note that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations",
+ "description": "Sonar is lightweight, affordable, fast, and simple to use — now featuring citations and the ability to customize sources. It is designed for companies seeking to integrate lightweight question-and-answer features optimized for speed.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "image",
"text",
- "file"
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "GPT",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 100000,
- "is_moderated": true
+ "context_length": 127072,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
"max_tokens",
- "reasoning",
- "response_format",
- "seed",
- "structured_outputs",
- "tool_choice",
- "tools"
+ "presence_penalty",
+ "temperature",
+ "top_k",
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/o3-mini",
- "name": "OpenAI: o3 Mini",
+ "id": "perplexity/sonar-deep-research",
+ "name": "Perplexity: Sonar Deep Research",
"provider": "openrouter",
- "family": "openai",
- "created_at": "2025-01-31 20:28:41 +0100",
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "perplexity",
+ "created_at": "2025-03-07 12:34:06 +1100",
+ "context_window": 128000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -24910,21 +25373,19 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.1,
- "output_per_million": 4.4,
- "cached_input_per_million": 0.55
+ "input_per_million": 2.0,
+ "output_per_million": 8.0,
+ "reasoning_output_per_million": 3.0
}
}
},
"metadata": {
- "description": "OpenAI o3-mini is a cost-efficient language model optimized for STEM reasoning tasks, particularly excelling in science, mathematics, and coding.\n\nThis model supports the `reasoning_effort` parameter, which can be set to \"high\", \"medium\", or \"low\" to control the thinking time of the model. The default is \"medium\". OpenRouter also offers the model slug `openai/o3-mini-high` to default the parameter to \"high\".\n\nThe model features three adjustable reasoning effort levels and supports key developer capabilities including function calling, structured outputs, and streaming, though it does not include vision processing capabilities.\n\nThe model demonstrates significant improvements over its predecessor, with expert testers preferring its responses 56% of the time and noting a 39% reduction in major errors on complex questions. With medium reasoning effort settings, o3-mini matches the performance of the larger o1 model on challenging reasoning evaluations like AIME and GPQA, while maintaining lower latency and cost.",
+ "description": "Sonar Deep Research is a research-focused model designed for multi-step retrieval, synthesis, and reasoning across complex topics. It autonomously searches, reads, and evaluates sources, refining its approach as it gathers information. This enables comprehensive report generation across domains like finance, technology, health, and current events.\n\nNotes on Pricing ([Source](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-deep-research)) \n- Input tokens comprise of Prompt tokens (user prompt) + Citation tokens (these are processed tokens from running searches)\n- Deep Research runs multiple searches to conduct exhaustive research. Searches are priced at $5/1000 searches. A request that does 30 searches will cost $0.15 in this step.\n- Reasoning is a distinct step in Deep Research since it does extensive automated reasoning through all the material it gathers during its research phase. Reasoning tokens here are a bit different than the CoTs in the answer - these are tokens that we use to reason through the research material prior to generating the outputs via the CoTs. Reasoning tokens are priced at $3/1M tokens",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -24934,61 +25395,63 @@
"text"
],
"tokenizer": "Other",
- "instruct_type": null
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 100000,
- "is_moderated": true
+ "context_length": 128000,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
"max_tokens",
- "response_format",
- "seed",
- "structured_outputs",
- "tool_choice",
- "tools"
+ "presence_penalty",
+ "reasoning",
+ "temperature",
+ "top_k",
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/o3-mini-high",
- "name": "OpenAI: o3 Mini High",
+ "id": "perplexity/sonar-pro",
+ "name": "Perplexity: Sonar Pro",
"provider": "openrouter",
- "family": "openai",
- "created_at": "2025-02-12 16:03:31 +0100",
+ "family": "perplexity",
+ "created_at": "2025-03-07 12:53:43 +1100",
"context_window": 200000,
- "max_output_tokens": 100000,
+ "max_output_tokens": 8000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.1,
- "output_per_million": 4.4,
- "cached_input_per_million": 0.55
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
}
}
},
"metadata": {
- "description": "OpenAI o3-mini-high is the same model as [o3-mini](/openai/o3-mini) with reasoning_effort set to high. \n\no3-mini is a cost-efficient language model optimized for STEM reasoning tasks, particularly excelling in science, mathematics, and coding. The model features three adjustable reasoning effort levels and supports key developer capabilities including function calling, structured outputs, and streaming, though it does not include vision processing capabilities.\n\nThe model demonstrates significant improvements over its predecessor, with expert testers preferring its responses 56% of the time and noting a 39% reduction in major errors on complex questions. With medium reasoning effort settings, o3-mini matches the performance of the larger o1 model on challenging reasoning evaluations like AIME and GPQA, while maintaining lower latency and cost.",
+ "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nFor enterprises seeking more advanced capabilities, the Sonar Pro API can handle in-depth, multi-step queries with added extensibility, like double the number of citations per search as Sonar on average. Plus, with a larger context window, it can handle longer and more nuanced searches and follow-up questions. ",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -24998,227 +25461,215 @@
},
"top_provider": {
"context_length": 200000,
- "max_completion_tokens": 100000,
- "is_moderated": true
+ "max_completion_tokens": 8000,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"max_tokens",
- "response_format",
- "seed",
- "structured_outputs",
- "tool_choice",
- "tools"
+ "presence_penalty",
+ "temperature",
+ "top_k",
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/o3-pro",
- "name": "OpenAI: o3 Pro",
+ "id": "perplexity/sonar-reasoning",
+ "name": "Perplexity: Sonar Reasoning",
"provider": "openrouter",
- "family": "openai",
- "created_at": "2025-06-11 01:32:32 +0200",
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "perplexity",
+ "created_at": "2025-01-29 17:11:47 +1100",
+ "context_window": 127000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "file",
- "image"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 20.0,
- "output_per_million": 80.0
+ "input_per_million": 1.0,
+ "output_per_million": 5.0
}
}
},
"metadata": {
- "description": "The o-series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o3-pro model uses more compute to think harder and provide consistently better answers.\n\nNote that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations",
+ "description": "Sonar Reasoning is a reasoning model provided by Perplexity based on [DeepSeek R1](/deepseek/deepseek-r1).\n\nIt allows developers to utilize long chain of thought with built-in web search. Sonar Reasoning is uncensored and hosted in US datacenters. ",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "file",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "GPT",
- "instruct_type": null
+ "tokenizer": "Other",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 100000,
- "is_moderated": true
+ "context_length": 127000,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"include_reasoning",
"max_tokens",
+ "presence_penalty",
"reasoning",
- "response_format",
- "seed",
- "structured_outputs",
- "tool_choice",
- "tools"
+ "temperature",
+ "top_k",
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/o4-mini",
- "name": "OpenAI: o4 Mini",
+ "id": "perplexity/sonar-reasoning-pro",
+ "name": "Perplexity: Sonar Reasoning Pro",
"provider": "openrouter",
- "family": "openai",
- "created_at": "2025-04-16 18:29:02 +0200",
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "perplexity",
+ "created_at": "2025-03-07 13:08:28 +1100",
+ "context_window": 128000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.1,
- "output_per_million": 4.4,
- "cached_input_per_million": 0.275
+ "input_per_million": 2.0,
+ "output_per_million": 8.0
}
}
},
"metadata": {
- "description": "OpenAI o4-mini is a compact reasoning model in the o-series, optimized for fast, cost-efficient performance while retaining strong multimodal and agentic capabilities. It supports tool use and demonstrates competitive reasoning and coding performance across benchmarks like AIME (99.5% with Python) and SWE-bench, outperforming its predecessor o3-mini and even approaching o3 in some domains.\n\nDespite its smaller size, o4-mini exhibits high accuracy in STEM tasks, visual problem solving (e.g., MathVista, MMMU), and code editing. It is especially well-suited for high-throughput scenarios where latency or cost is critical. Thanks to its efficient architecture and refined reinforcement learning training, o4-mini can chain tools, generate structured outputs, and solve multi-step tasks with minimal delay—often in under a minute.",
+ "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nSonar Reasoning Pro is a premier reasoning model powered by DeepSeek R1 with Chain of Thought (CoT). Designed for advanced use cases, it supports in-depth, multi-step queries with a larger context window and can surface more citations per search, enabling more comprehensive and extensible responses.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "image",
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Other",
- "instruct_type": null
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 100000,
- "is_moderated": true
+ "context_length": 128000,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"include_reasoning",
"max_tokens",
+ "presence_penalty",
"reasoning",
- "response_format",
- "seed",
- "structured_outputs",
- "tool_choice",
- "tools"
+ "temperature",
+ "top_k",
+ "top_p",
+ "web_search_options"
]
}
},
{
- "id": "openai/o4-mini-high",
- "name": "OpenAI: o4 Mini High",
+ "id": "pygmalionai/mythalion-13b",
+ "name": "Pygmalion: Mythalion 13B",
"provider": "openrouter",
- "family": "openai",
- "created_at": "2025-04-16 19:23:32 +0200",
- "context_window": 200000,
- "max_output_tokens": 100000,
+ "family": "pygmalionai",
+ "created_at": "2023-09-02 10:00:00 +1000",
+ "context_window": 4096,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
- "text",
- "file"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.1,
- "output_per_million": 4.4,
- "cached_input_per_million": 0.275
+ "input_per_million": 0.7,
+ "output_per_million": 1.1
}
}
},
"metadata": {
- "description": "OpenAI o4-mini-high is the same model as [o4-mini](/openai/o4-mini) with reasoning_effort set to high. \n\nOpenAI o4-mini is a compact reasoning model in the o-series, optimized for fast, cost-efficient performance while retaining strong multimodal and agentic capabilities. It supports tool use and demonstrates competitive reasoning and coding performance across benchmarks like AIME (99.5% with Python) and SWE-bench, outperforming its predecessor o3-mini and even approaching o3 in some domains.\n\nDespite its smaller size, o4-mini exhibits high accuracy in STEM tasks, visual problem solving (e.g., MathVista, MMMU), and code editing. It is especially well-suited for high-throughput scenarios where latency or cost is critical. Thanks to its efficient architecture and refined reinforcement learning training, o4-mini can chain tools, generate structured outputs, and solve multi-step tasks with minimal delay—often in under a minute.",
+ "description": "A blend of the new Pygmalion-13b and MythoMax. #merge",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
- "text",
- "file"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama2",
+ "instruct_type": "alpaca"
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 100000,
- "is_moderated": true
+ "context_length": 4096,
+ "max_completion_tokens": 4096,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
"max_tokens",
- "reasoning",
- "response_format",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
"seed",
- "structured_outputs",
- "tool_choice",
- "tools"
+ "stop",
+ "temperature",
+ "top_k",
+ "top_p"
]
}
},
{
- "id": "opengvlab/internvl3-14b",
- "name": "OpenGVLab: InternVL3 14B",
+ "id": "qwen/qwen-2-72b-instruct",
+ "name": "Qwen 2 72B Instruct",
"provider": "openrouter",
- "family": "opengvlab",
- "created_at": "2025-04-30 15:55:55 +0200",
- "context_window": 12288,
- "max_output_tokens": null,
+ "family": "qwen",
+ "created_at": "2024-06-07 10:00:00 +1000",
+ "context_window": 32768,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -25226,50 +25677,57 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.39999999999999997
+ "input_per_million": 0.8999999999999999,
+ "output_per_million": 0.8999999999999999
}
}
},
"metadata": {
- "description": "The 14b version of the InternVL3 series. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.",
+ "description": "Qwen2 72B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 12288,
- "max_completion_tokens": null,
+ "context_length": 32768,
+ "max_completion_tokens": 4096,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "stop",
"temperature",
+ "top_k",
"top_p"
]
}
},
{
- "id": "openrouter/auto",
- "name": "Auto Router",
+ "id": "qwen/qwen-2.5-72b-instruct",
+ "name": "Qwen2.5 72B Instruct",
"provider": "openrouter",
- "family": "openrouter",
- "created_at": "2023-11-08 01:00:00 +0100",
- "context_window": 2000000,
+ "family": "qwen",
+ "created_at": "2024-09-19 10:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -25281,11 +25739,21 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.051830799999999996,
+ "output_per_million": 0.207424
+ }
+ }
+ },
"metadata": {
- "description": "Your prompt will be processed by a meta-model and routed to one of dozens of models (see below), optimizing for the best possible output.\n\nTo see which model was used, visit [Activity](/activity), or read the `model` attribute of the response. Your response will be priced at the same rate as the routed model.\n\nThe meta-model is powered by [Not Diamond](https://docs.notdiamond.ai/docs/how-not-diamond-works). Learn more in our [docs](/docs/model-routing).\n\nRequests will be routed to the following models:\n- [openai/gpt-4o-2024-08-06](/openai/gpt-4o-2024-08-06)\n- [openai/gpt-4o-2024-05-13](/openai/gpt-4o-2024-05-13)\n- [openai/gpt-4o-mini-2024-07-18](/openai/gpt-4o-mini-2024-07-18)\n- [openai/chatgpt-4o-latest](/openai/chatgpt-4o-latest)\n- [openai/o1-preview-2024-09-12](/openai/o1-preview-2024-09-12)\n- [openai/o1-mini-2024-09-12](/openai/o1-mini-2024-09-12)\n- [anthropic/claude-3.5-sonnet](/anthropic/claude-3.5-sonnet)\n- [anthropic/claude-3.5-haiku](/anthropic/claude-3.5-haiku)\n- [anthropic/claude-3-opus](/anthropic/claude-3-opus)\n- [anthropic/claude-2.1](/anthropic/claude-2.1)\n- [google/gemini-pro-1.5](/google/gemini-pro-1.5)\n- [google/gemini-flash-1.5](/google/gemini-flash-1.5)\n- [mistralai/mistral-large-2407](/mistralai/mistral-large-2407)\n- [mistralai/mistral-nemo](/mistralai/mistral-nemo)\n- [deepseek/deepseek-r1](/deepseek/deepseek-r1)\n- [meta-llama/llama-3.1-70b-instruct](/meta-llama/llama-3.1-70b-instruct)\n- [meta-llama/llama-3.1-405b-instruct](/meta-llama/llama-3.1-405b-instruct)\n- [mistralai/mixtral-8x22b-instruct](/mistralai/mixtral-8x22b-instruct)\n- [cohere/command-r-plus](/cohere/command-r-plus)\n- [cohere/command-r](/cohere/command-r)",
+ "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -25294,25 +25762,42 @@
"output_modalities": [
"text"
],
- "tokenizer": "Router",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": null,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
- "supported_parameters": []
+ "supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
+ "max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "response_format",
+ "seed",
+ "stop",
+ "temperature",
+ "tool_choice",
+ "tools",
+ "top_k",
+ "top_logprobs",
+ "top_p"
+ ]
}
},
{
- "id": "perplexity/r1-1776",
- "name": "Perplexity: R1 1776",
+ "id": "qwen/qwen-2.5-72b-instruct:free",
+ "name": "Qwen2.5 72B Instruct (free)",
"provider": "openrouter",
- "family": "perplexity",
- "created_at": "2025-02-19 23:42:09 +0100",
- "context_window": 128000,
+ "family": "qwen",
+ "created_at": "2024-09-19 10:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -25327,16 +25812,9 @@
"streaming",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 2.0,
- "output_per_million": 8.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "R1 1776 is a version of DeepSeek-R1 that has been post-trained to remove censorship constraints related to topics restricted by the Chinese government. The model retains its original reasoning capabilities while providing direct responses to a wider range of queries. R1 1776 is an offline chat model that does not use the perplexity search subsystem.\n\nThe model was tested on a multilingual dataset of over 1,000 examples covering sensitive topics to measure its likelihood of refusal or overly filtered responses. [Evaluation Results](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/GiN2VqC5hawUgAGJ6oHla.png) Its performance on math and reasoning benchmarks remains similar to the base R1 model. [Reasoning Performance](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/n4Z9Byqp2S7sKUvCvI40R.png)\n\nRead more on the [Blog Post](https://perplexity.ai/hub/blog/open-sourcing-r1-1776)",
+ "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -25345,98 +25823,105 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Qwen",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
+ "seed",
"stop",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "perplexity/sonar",
- "name": "Perplexity: Sonar",
+ "id": "qwen/qwen-2.5-7b-instruct",
+ "name": "Qwen2.5 7B Instruct",
"provider": "openrouter",
- "family": "perplexity",
- "created_at": "2025-01-27 22:36:48 +0100",
- "context_window": 127072,
+ "family": "qwen",
+ "created_at": "2024-10-16 11:00:00 +1100",
+ "context_window": 65536,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.0,
- "output_per_million": 1.0
+ "input_per_million": 0.04,
+ "output_per_million": 0.09999999999999999
}
}
},
"metadata": {
- "description": "Sonar is lightweight, affordable, fast, and simple to use — now featuring citations and the ability to customize sources. It is designed for companies seeking to integrate lightweight question-and-answer features optimized for speed.",
+ "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 127072,
+ "context_length": 65536,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
"temperature",
"top_k",
- "top_p",
- "web_search_options"
+ "top_p"
]
}
},
{
- "id": "perplexity/sonar-deep-research",
- "name": "Perplexity: Sonar Deep Research",
+ "id": "qwen/qwen-2.5-coder-32b-instruct",
+ "name": "Qwen2.5 Coder 32B Instruct",
"provider": "openrouter",
- "family": "perplexity",
- "created_at": "2025-03-07 02:34:06 +0100",
- "context_window": 128000,
+ "family": "qwen",
+ "created_at": "2024-11-12 10:40:00 +1100",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -25448,19 +25933,20 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 8.0,
- "reasoning_output_per_million": 3.0
+ "input_per_million": 0.0499797,
+ "output_per_million": 0.200016
}
}
},
"metadata": {
- "description": "Sonar Deep Research is a research-focused model designed for multi-step retrieval, synthesis, and reasoning across complex topics. It autonomously searches, reads, and evaluates sources, refining its approach as it gathers information. This enables comprehensive report generation across domains like finance, technology, health, and current events.\n\nNotes on Pricing ([Source](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-deep-research)) \n- Input tokens comprise of Prompt tokens (user prompt) + Citation tokens (these are processed tokens from running searches)\n- Deep Research runs multiple searches to conduct exhaustive research. Searches are priced at $5/1000 searches. A request that does 30 searches will cost $0.15 in this step.\n- Reasoning is a distinct step in Deep Research since it does extensive automated reasoning through all the material it gathers during its research phase. Reasoning tokens here are a bit different than the CoTs in the answer - these are tokens that we use to reason through the research material prior to generating the outputs via the CoTs. Reasoning tokens are priced at $3/1M tokens",
+ "description": "Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:\n\n- Significantly improvements in **code generation**, **code reasoning** and **code fixing**. \n- A more comprehensive foundation for real-world applications such as **Code Agents**. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.\n\nTo read more about its evaluation results, check out [Qwen 2.5 Coder's blog](https://qwenlm.github.io/blog/qwen2.5-coder-family/).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -25469,218 +25955,231 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Qwen",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 128000,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "reasoning",
+ "repetition_penalty",
+ "response_format",
+ "seed",
+ "stop",
"temperature",
"top_k",
- "top_p",
- "web_search_options"
+ "top_logprobs",
+ "top_p"
]
}
},
{
- "id": "perplexity/sonar-pro",
- "name": "Perplexity: Sonar Pro",
+ "id": "qwen/qwen-2.5-coder-32b-instruct:free",
+ "name": "Qwen2.5 Coder 32B Instruct (free)",
"provider": "openrouter",
- "family": "perplexity",
- "created_at": "2025-03-07 02:53:43 +0100",
- "context_window": 200000,
- "max_output_tokens": 8000,
+ "family": "qwen",
+ "created_at": "2024-11-12 10:40:00 +1100",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nFor enterprises seeking more advanced capabilities, the Sonar Pro API can handle in-depth, multi-step queries with added extensibility, like double the number of citations per search as Sonar on average. Plus, with a larger context window, it can handle longer and more nuanced searches and follow-up questions. ",
+ "description": "Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:\n\n- Significantly improvements in **code generation**, **code reasoning** and **code fixing**. \n- A more comprehensive foundation for real-world applications such as **Code Agents**. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.\n\nTo read more about its evaluation results, check out [Qwen 2.5 Coder's blog](https://qwenlm.github.io/blog/qwen2.5-coder-family/).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 200000,
- "max_completion_tokens": 8000,
+ "context_length": 32768,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
+ "seed",
+ "stop",
"temperature",
"top_k",
- "top_p",
- "web_search_options"
+ "top_logprobs",
+ "top_p"
]
}
},
{
- "id": "perplexity/sonar-reasoning",
- "name": "Perplexity: Sonar Reasoning",
+ "id": "qwen/qwen-2.5-vl-7b-instruct",
+ "name": "Qwen: Qwen2.5-VL 7B Instruct",
"provider": "openrouter",
- "family": "perplexity",
- "created_at": "2025-01-29 07:11:47 +0100",
- "context_window": 127000,
+ "family": "qwen",
+ "created_at": "2024-08-28 10:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.0,
- "output_per_million": 5.0
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.19999999999999998
}
}
},
"metadata": {
- "description": "Sonar Reasoning is a reasoning model provided by Perplexity based on [DeepSeek R1](/deepseek/deepseek-r1).\n\nIt allows developers to utilize long chain of thought with built-in web search. Sonar Reasoning is uncensored and hosted in US datacenters. ",
+ "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Qwen",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 127000,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "reasoning",
+ "repetition_penalty",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
"temperature",
"top_k",
- "top_p",
- "web_search_options"
+ "top_logprobs",
+ "top_p"
]
}
},
{
- "id": "perplexity/sonar-reasoning-pro",
- "name": "Perplexity: Sonar Reasoning Pro",
+ "id": "qwen/qwen-max",
+ "name": "Qwen: Qwen-Max ",
"provider": "openrouter",
- "family": "perplexity",
- "created_at": "2025-03-07 03:08:28 +0100",
- "context_window": 128000,
- "max_output_tokens": null,
+ "family": "qwen",
+ "created_at": "2025-02-01 20:31:29 +1100",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 8.0
+ "input_per_million": 1.5999999999999999,
+ "output_per_million": 6.3999999999999995,
+ "cached_input_per_million": 0.64
}
}
},
"metadata": {
- "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nSonar Reasoning Pro is a premier reasoning model powered by DeepSeek R1 with Chain of Thought (CoT). Designed for advanced use cases, it supports in-depth, multi-step queries with a larger context window and can surface more citations per search, enabling more comprehensive and extensible responses.",
+ "description": "Qwen-Max, based on Qwen2.5, provides the best inference performance among [Qwen models](/qwen), especially for complex multi-step tasks. It's a large-scale MoE model that has been pretrained on over 20 trillion tokens and further post-trained with curated Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF) methodologies. The parameter count is unknown.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Qwen",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": null,
+ "context_length": 32768,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
"max_tokens",
"presence_penalty",
- "reasoning",
+ "response_format",
+ "seed",
"temperature",
- "top_k",
- "top_p",
- "web_search_options"
+ "tool_choice",
+ "tools",
+ "top_p"
]
}
},
{
- "id": "pygmalionai/mythalion-13b",
- "name": "Pygmalion: Mythalion 13B",
+ "id": "qwen/qwen-plus",
+ "name": "Qwen: Qwen-Plus",
"provider": "openrouter",
- "family": "pygmalionai",
- "created_at": "2023-09-02 02:00:00 +0200",
- "context_window": 4096,
- "max_output_tokens": 4096,
+ "family": "qwen",
+ "created_at": "2025-02-01 22:37:20 +1100",
+ "context_window": 131072,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -25691,18 +26190,21 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7,
- "output_per_million": 1.1
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 1.2,
+ "cached_input_per_million": 0.16
}
}
},
"metadata": {
- "description": "A blend of the new Pygmalion-13b and MythoMax. #merge",
+ "description": "Qwen-Plus, based on the Qwen2.5 foundation model, is a 131K context model with a balanced performance, speed, and cost combination.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -25711,37 +26213,35 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama2",
- "instruct_type": "alpaca"
+ "tokenizer": "Qwen",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 4096,
- "max_completion_tokens": 4096,
+ "context_length": 131072,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
- "stop",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "qwen/qwen-2-72b-instruct",
- "name": "Qwen 2 72B Instruct",
+ "id": "qwen/qwen-turbo",
+ "name": "Qwen: Qwen-Turbo",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2024-06-07 02:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": 4096,
+ "created_at": "2025-02-01 22:56:14 +1100",
+ "context_window": 1000000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -25753,18 +26253,20 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.8999999999999999,
- "output_per_million": 0.8999999999999999
+ "input_per_million": 0.049999999999999996,
+ "output_per_million": 0.19999999999999998,
+ "cached_input_per_million": 0.02
}
}
},
"metadata": {
- "description": "Qwen2 72B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
+ "description": "Qwen-Turbo, based on Qwen2.5, is a 1M context model that provides fast speed and low cost, suitable for simple tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -25774,40 +26276,39 @@
"text"
],
"tokenizer": "Qwen",
- "instruct_type": "chatml"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 4096,
+ "context_length": 1000000,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
- "stop",
+ "response_format",
+ "seed",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "qwen/qwen-2.5-72b-instruct",
- "name": "Qwen2.5 72B Instruct",
+ "id": "qwen/qwen-vl-max",
+ "name": "Qwen: Qwen VL Max",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2024-09-19 02:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": null,
+ "created_at": "2025-02-02 05:25:04 +1100",
+ "context_window": 7500,
+ "max_output_tokens": 1500,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -25815,69 +26316,59 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.0666396,
- "output_per_million": 0.26668800000000004
+ "input_per_million": 0.7999999999999999,
+ "output_per_million": 3.1999999999999997
}
}
},
"metadata": {
- "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
+ "description": "Qwen VL Max is a visual understanding model with 7500 tokens context length. It excels in delivering optimal performance for a broader spectrum of complex tasks.\n",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Qwen",
- "instruct_type": "chatml"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 7500,
+ "max_completion_tokens": 1500,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
- "stop",
"temperature",
- "tool_choice",
- "tools",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen-2.5-72b-instruct:free",
- "name": "Qwen2.5 72B Instruct (free)",
+ "id": "qwen/qwen-vl-plus",
+ "name": "Qwen: Qwen VL Plus",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2024-09-19 02:00:00 +0200",
- "context_window": 32768,
- "max_output_tokens": null,
+ "created_at": "2025-02-05 15:54:15 +1100",
+ "context_window": 7500,
+ "max_output_tokens": 1500,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -25885,57 +26376,59 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.21,
+ "output_per_million": 0.63
+ }
+ }
+ },
"metadata": {
- "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
+ "description": "Qwen's Enhanced Large Visual Language Model. Significantly upgraded for detailed recognition capabilities and text recognition abilities, supporting ultra-high pixel resolutions up to millions of pixels and extreme aspect ratios for image input. It delivers significant performance across a broad range of visual tasks.\n",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Qwen",
- "instruct_type": "chatml"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 7500,
+ "max_completion_tokens": 1500,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
- "stop",
"temperature",
- "top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen-2.5-7b-instruct",
- "name": "Qwen2.5 7B Instruct",
+ "id": "qwen/qwen2.5-vl-32b-instruct",
+ "name": "Qwen: Qwen2.5 VL 32B Instruct",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2024-10-16 02:00:00 +0200",
- "context_window": 65536,
+ "created_at": "2025-03-25 05:10:38 +1100",
+ "context_window": 16384,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -25949,26 +26442,27 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.04,
- "output_per_million": 0.09999999999999999
+ "input_per_million": 0.01999188,
+ "output_per_million": 0.0800064
}
}
},
"metadata": {
- "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
+ "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Qwen",
- "instruct_type": "chatml"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 65536,
+ "context_length": 16384,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -25976,6 +26470,7 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -25986,22 +26481,24 @@
"structured_outputs",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen-2.5-coder-32b-instruct",
- "name": "Qwen2.5 Coder 32B Instruct",
+ "id": "qwen/qwen2.5-vl-32b-instruct:free",
+ "name": "Qwen: Qwen2.5 VL 32B Instruct (free)",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2024-11-12 00:40:00 +0100",
- "context_window": 32768,
+ "created_at": "2025-03-25 05:10:38 +1100",
+ "context_window": 8192,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -26012,29 +26509,23 @@
"structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.0499797,
- "output_per_million": 0.200016
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:\n\n- Significantly improvements in **code generation**, **code reasoning** and **code fixing**. \n- A more comprehensive foundation for real-world applications such as **Code Agents**. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.\n\nTo read more about its evaluation results, check out [Qwen 2.5 Coder's blog](https://qwenlm.github.io/blog/qwen2.5-coder-family/).",
+ "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Qwen",
- "instruct_type": "chatml"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 8192,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -26058,17 +26549,18 @@
}
},
{
- "id": "qwen/qwen-2.5-coder-32b-instruct:free",
- "name": "Qwen2.5 Coder 32B Instruct (free)",
+ "id": "qwen/qwen2.5-vl-72b-instruct",
+ "name": "Qwen: Qwen2.5 VL 72B Instruct",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2024-11-12 00:40:00 +0100",
+ "created_at": "2025-02-01 22:45:11 +1100",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -26078,19 +26570,27 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.0999594,
+ "output_per_million": 0.400032
+ }
+ }
+ },
"metadata": {
- "description": "Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:\n\n- Significantly improvements in **code generation**, **code reasoning** and **code fixing**. \n- A more comprehensive foundation for real-world applications such as **Code Agents**. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.\n\nTo read more about its evaluation results, check out [Qwen 2.5 Coder's blog](https://qwenlm.github.io/blog/qwen2.5-coder-family/).",
+ "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
"tokenizer": "Qwen",
- "instruct_type": "chatml"
+ "instruct_type": null
},
"top_provider": {
"context_length": 32768,
@@ -26116,11 +26616,11 @@
}
},
{
- "id": "qwen/qwen-2.5-vl-7b-instruct",
- "name": "Qwen: Qwen2.5-VL 7B Instruct",
+ "id": "qwen/qwen2.5-vl-72b-instruct:free",
+ "name": "Qwen: Qwen2.5 VL 72B Instruct (free)",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2024-08-28 02:00:00 +0200",
+ "created_at": "2025-02-01 22:45:11 +1100",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -26135,19 +26635,11 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.19999999999999998
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).",
+ "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -26168,31 +26660,26 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen-max",
- "name": "Qwen: Qwen-Max ",
+ "id": "qwen/qwen3-14b",
+ "name": "Qwen: Qwen3 14B",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-02-01 10:31:29 +0100",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "created_at": "2025-04-29 07:41:18 +1000",
+ "context_window": 40960,
+ "max_output_tokens": 40960,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -26205,19 +26692,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.5999999999999999,
- "output_per_million": 6.3999999999999995,
- "cached_input_per_million": 0.64
+ "input_per_million": 0.06,
+ "output_per_million": 0.24
}
}
},
"metadata": {
- "description": "Qwen-Max, based on Qwen2.5, provides the best inference performance among [Qwen models](/qwen), especially for complex multi-step tasks. It's a large-scale MoE model that has been pretrained on over 20 trillion tokens and further post-trained with curated Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF) methodologies. The parameter count is unknown.",
+ "description": "Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, programming, and logical inference, and a \"non-thinking\" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -26226,35 +26713,46 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": null
+ "tokenizer": "Qwen3",
+ "instruct_type": "qwen3"
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 8192,
+ "context_length": 40960,
+ "max_completion_tokens": 40960,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
+ "stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen-plus",
- "name": "Qwen: Qwen-Plus",
+ "id": "qwen/qwen3-14b:free",
+ "name": "Qwen: Qwen3 14B (free)",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-02-01 12:37:20 +0100",
- "context_window": 131072,
- "max_output_tokens": 8192,
+ "created_at": "2025-04-29 07:41:18 +1000",
+ "context_window": 40960,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -26266,20 +26764,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 1.2,
- "cached_input_per_million": 0.16
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Qwen-Plus, based on the Qwen2.5 foundation model, is a 131K context model with a balanced performance, speed, and cost combination.",
+ "description": "Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, programming, and logical inference, and a \"non-thinking\" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -26288,35 +26777,42 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": null
+ "tokenizer": "Qwen3",
+ "instruct_type": "qwen3"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 8192,
+ "context_length": 40960,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "reasoning",
+ "repetition_penalty",
"seed",
+ "stop",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen-turbo",
- "name": "Qwen: Qwen-Turbo",
+ "id": "qwen/qwen3-235b-a22b",
+ "name": "Qwen: Qwen3 235B A22B",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-02-01 12:56:14 +0100",
- "context_window": 1000000,
- "max_output_tokens": 8192,
+ "created_at": "2025-04-29 07:29:17 +1000",
+ "context_window": 40960,
+ "max_output_tokens": 40960,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -26329,19 +26825,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.049999999999999996,
- "output_per_million": 0.19999999999999998,
- "cached_input_per_million": 0.02
+ "input_per_million": 0.13,
+ "output_per_million": 0.6
}
}
},
"metadata": {
- "description": "Qwen-Turbo, based on Qwen2.5, is a 1M context model that provides fast speed and low cost, suitable for simple tasks.",
+ "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -26350,40 +26846,50 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": null
+ "tokenizer": "Qwen3",
+ "instruct_type": "qwen3"
},
"top_provider": {
- "context_length": 1000000,
- "max_completion_tokens": 8192,
+ "context_length": 40960,
+ "max_completion_tokens": 40960,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
+ "stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen-vl-max",
- "name": "Qwen: Qwen VL Max",
+ "id": "qwen/qwen3-235b-a22b-2507",
+ "name": "Qwen: Qwen3 235B A22B Instruct 2507",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-02-01 19:25:04 +0100",
- "context_window": 7500,
- "max_output_tokens": 1500,
+ "created_at": "2025-07-22 03:39:15 +1000",
+ "context_window": 262144,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -26391,59 +26897,70 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7999999999999999,
- "output_per_million": 3.1999999999999997
+ "input_per_million": 0.077968332,
+ "output_per_million": 0.31202496
}
}
},
"metadata": {
- "description": "Qwen VL Max is a visual understanding model with 7500 tokens context length. It excels in delivering optimal performance for a broader spectrum of complex tasks.\n",
+ "description": "Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following, logical reasoning, math, code, and tool usage. The model supports a native 262K context length and does not implement \"thinking mode\" ( blocks).\n\nCompared to its base variant, this version delivers significant gains in knowledge coverage, long-context reasoning, coding benchmarks, and alignment with open-ended tasks. It is particularly strong on multilingual understanding, math reasoning (e.g., AIME, HMMT), and alignment evaluations like Arena-Hard and WritingBench.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
+ "tokenizer": "Qwen3",
"instruct_type": null
},
"top_provider": {
- "context_length": 7500,
- "max_completion_tokens": 1500,
+ "context_length": 262144,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
+ "stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen-vl-plus",
- "name": "Qwen: Qwen VL Plus",
+ "id": "qwen/qwen3-235b-a22b-thinking-2507",
+ "name": "Qwen: Qwen3 235B A22B Thinking 2507",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-02-05 05:54:15 +0100",
- "context_window": 7500,
- "max_output_tokens": 1500,
+ "created_at": "2025-07-25 23:19:17 +1000",
+ "context_window": 262144,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -26451,59 +26968,71 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.21,
- "output_per_million": 0.63
+ "input_per_million": 0.077968332,
+ "output_per_million": 0.31202496
}
}
},
"metadata": {
- "description": "Qwen's Enhanced Large Visual Language Model. Significantly upgraded for detailed recognition capabilities and text recognition abilities, supporting ultra-high pixel resolutions up to millions of pixels and extreme aspect ratios for image input. It delivers significant performance across a broad range of visual tasks.\n",
+ "description": "Qwen3-235B-A22B-Thinking-2507 is a high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It activates 22B of its 235B parameters per forward pass and natively supports up to 262,144 tokens of context. This \"thinking-only\" variant enhances structured logical reasoning, mathematics, science, and long-form generation, showing strong benchmark performance across AIME, SuperGPQA, LiveCodeBench, and MMLU-Redux. It enforces a special reasoning mode () and is designed for high-token outputs (up to 81,920 tokens) in challenging domains.\n\nThe model is instruction-tuned and excels at step-by-step reasoning, tool use, agentic workflows, and multilingual tasks. This release represents the most capable open-source variant in the Qwen3-235B series, surpassing many closed models in structured reasoning use cases.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": null
+ "tokenizer": "Qwen3",
+ "instruct_type": "qwen3"
},
"top_provider": {
- "context_length": 7500,
- "max_completion_tokens": 1500,
+ "context_length": 262144,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
+ "stop",
"temperature",
+ "tool_choice",
+ "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen2.5-vl-32b-instruct",
- "name": "Qwen: Qwen2.5 VL 32B Instruct",
+ "id": "qwen/qwen3-235b-a22b:free",
+ "name": "Qwen: Qwen3 235B A22B (free)",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-03-24 19:10:38 +0100",
- "context_window": 16384,
+ "created_at": "2025-04-29 07:29:17 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -26511,50 +27040,47 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.01999188,
- "output_per_million": 0.0800064
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.",
+ "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": null
+ "tokenizer": "Qwen3",
+ "instruct_type": "qwen3"
},
"top_provider": {
- "context_length": 16384,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -26562,18 +27088,17 @@
}
},
{
- "id": "qwen/qwen2.5-vl-32b-instruct:free",
- "name": "Qwen: Qwen2.5 VL 32B Instruct (free)",
+ "id": "qwen/qwen3-30b-a3b",
+ "name": "Qwen: Qwen3 30B A3B",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-03-24 19:10:38 +0100",
- "context_window": 8192,
+ "created_at": "2025-04-29 08:16:44 +1000",
+ "context_window": 40960,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -26581,42 +27106,54 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.01999188,
+ "output_per_million": 0.0800064
+ }
+ }
+ },
"metadata": {
- "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.",
+ "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": null
+ "tokenizer": "Qwen3",
+ "instruct_type": "qwen3"
},
"top_provider": {
- "context_length": 8192,
+ "context_length": 40960,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -26624,18 +27161,17 @@
}
},
{
- "id": "qwen/qwen2.5-vl-72b-instruct",
- "name": "Qwen: Qwen2.5 VL 72B Instruct",
+ "id": "qwen/qwen3-30b-a3b-instruct-2507",
+ "name": "Qwen: Qwen3 30B A3B Instruct 2507",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-02-01 12:45:11 +0100",
- "context_window": 32768,
+ "created_at": "2025-07-30 02:36:05 +1000",
+ "context_window": 262144,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -26643,32 +27179,33 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.0999594,
- "output_per_million": 0.400032
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.",
+ "description": "Qwen3-30B-A3B-Instruct-2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference. It operates in non-thinking mode and is designed for high-quality instruction following, multilingual understanding, and agentic tool use. Post-trained on instruction data, it demonstrates competitive performance across reasoning (AIME, ZebraLogic), coding (MultiPL-E, LiveCodeBench), and alignment (IFEval, WritingBench) benchmarks. It outperforms its non-instruct variant on subjective and open-ended tasks while retaining strong factual and coding performance.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
+ "tokenizer": "Qwen3",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 262144,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -26678,12 +27215,13 @@
"logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -26691,18 +27229,17 @@
}
},
{
- "id": "qwen/qwen2.5-vl-72b-instruct:free",
- "name": "Qwen: Qwen2.5 VL 72B Instruct (free)",
+ "id": "qwen/qwen3-30b-a3b:free",
+ "name": "Qwen: Qwen3 30B A3B (free)",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-02-01 12:45:11 +0100",
- "context_window": 32768,
+ "created_at": "2025-04-29 08:16:44 +1000",
+ "context_window": 40960,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -26710,51 +27247,55 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {},
"metadata": {
- "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.",
+ "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": null
+ "tokenizer": "Qwen3",
+ "instruct_type": "qwen3"
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 40960,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "reasoning",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen3-14b",
- "name": "Qwen: Qwen3 14B",
+ "id": "qwen/qwen3-32b",
+ "name": "Qwen: Qwen3 32B",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-04-28 23:41:18 +0200",
+ "created_at": "2025-04-29 07:32:25 +1000",
"context_window": 40960,
- "max_output_tokens": 40960,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -26773,13 +27314,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.06,
- "output_per_million": 0.24
+ "input_per_million": 0.017992691999999998,
+ "output_per_million": 0.07200576
}
}
},
"metadata": {
- "description": "Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, programming, and logical inference, and a \"non-thinking\" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.",
+ "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -26793,7 +27334,7 @@
},
"top_provider": {
"context_length": 40960,
- "max_completion_tokens": 40960,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -26821,11 +27362,11 @@
}
},
{
- "id": "qwen/qwen3-14b:free",
- "name": "Qwen: Qwen3 14B (free)",
+ "id": "qwen/qwen3-4b:free",
+ "name": "Qwen: Qwen3 4B (free)",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-04-28 23:41:18 +0200",
+ "created_at": "2025-05-01 02:38:24 +1000",
"context_window": 40960,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -26839,11 +27380,12 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {},
"metadata": {
- "description": "Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, programming, and logical inference, and a \"non-thinking\" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.",
+ "description": "Qwen3-4B is a 4 billion parameter dense language model from the Qwen3 series, designed to support both general-purpose and reasoning-intensive tasks. It introduces a dual-mode architecture—thinking and non-thinking—allowing dynamic switching between high-precision logical reasoning and efficient dialogue generation. This makes it well-suited for multi-turn chat, instruction following, and complex agent workflows.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -26864,30 +27406,28 @@
"supported_parameters": [
"frequency_penalty",
"include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
"reasoning",
- "repetition_penalty",
- "seed",
+ "response_format",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen3-235b-a22b",
- "name": "Qwen: Qwen3 235B A22B",
+ "id": "qwen/qwen3-8b",
+ "name": "Qwen: Qwen3 8B",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-04-28 23:29:17 +0200",
- "context_window": 40960,
- "max_output_tokens": 40960,
+ "created_at": "2025-04-29 07:43:52 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 20000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -26899,20 +27439,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.13,
- "output_per_million": 0.6
+ "input_per_million": 0.035,
+ "output_per_million": 0.13799999999999998
}
}
},
"metadata": {
- "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.",
+ "description": "Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between \"thinking\" mode for math, coding, and logical inference, and \"non-thinking\" mode for general conversation. The model is fine-tuned for instruction-following, agent integration, creative writing, and multilingual use across 100+ languages and dialects. It natively supports a 32K token context window and can extend to 131K tokens with YaRN scaling.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -26925,8 +27463,8 @@
"instruct_type": "qwen3"
},
"top_provider": {
- "context_length": 40960,
- "max_completion_tokens": 40960,
+ "context_length": 128000,
+ "max_completion_tokens": 20000,
"is_moderated": false
},
"per_request_limits": null,
@@ -26934,33 +27472,27 @@
"frequency_penalty",
"include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen3-235b-a22b-2507",
- "name": "Qwen: Qwen3 235B A22B Instruct 2507",
+ "id": "qwen/qwen3-8b:free",
+ "name": "Qwen: Qwen3 8B (free)",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-07-21 19:39:15 +0200",
- "context_window": 262144,
- "max_output_tokens": null,
+ "created_at": "2025-04-29 07:43:52 +1000",
+ "context_window": 40960,
+ "max_output_tokens": 40960,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -26972,20 +27504,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.077968332,
- "output_per_million": 0.31202496
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following, logical reasoning, math, code, and tool usage. The model supports a native 262K context length and does not implement \"thinking mode\" ( blocks).\n\nCompared to its base variant, this version delivers significant gains in knowledge coverage, long-context reasoning, coding benchmarks, and alignment with open-ended tasks. It is particularly strong on multilingual understanding, math reasoning (e.g., AIME, HMMT), and alignment evaluations like Arena-Hard and WritingBench.",
+ "description": "Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between \"thinking\" mode for math, coding, and logical inference, and \"non-thinking\" mode for general conversation. The model is fine-tuned for instruction-following, agent integration, creative writing, and multilingual use across 100+ languages and dialects. It natively supports a 32K token context window and can extend to 131K tokens with YaRN scaling.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -26995,29 +27518,27 @@
"text"
],
"tokenizer": "Qwen3",
- "instruct_type": null
+ "instruct_type": "qwen3"
},
"top_provider": {
- "context_length": 262144,
- "max_completion_tokens": null,
+ "context_length": 40960,
+ "max_completion_tokens": 40960,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -27025,11 +27546,11 @@
}
},
{
- "id": "qwen/qwen3-235b-a22b-thinking-2507",
- "name": "Qwen: Qwen3 235B A22B Thinking 2507",
+ "id": "qwen/qwen3-coder",
+ "name": "Qwen: Qwen3 Coder ",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-07-25 15:19:17 +0200",
+ "created_at": "2025-07-23 10:29:06 +1000",
"context_window": 262144,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -27050,13 +27571,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.077968332,
- "output_per_million": 0.31202496
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.7999999999999999
}
}
},
"metadata": {
- "description": "Qwen3-235B-A22B-Thinking-2507 is a high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It activates 22B of its 235B parameters per forward pass and natively supports up to 262,144 tokens of context. This \"thinking-only\" variant enhances structured logical reasoning, mathematics, science, and long-form generation, showing strong benchmark performance across AIME, SuperGPQA, LiveCodeBench, and MMLU-Redux. It enforces a special reasoning mode () and is designed for high-token outputs (up to 81,920 tokens) in challenging domains.\n\nThe model is instruction-tuned and excels at step-by-step reasoning, tool use, agentic workflows, and multilingual tasks. This release represents the most capable open-source variant in the Qwen3-235B series, surpassing many closed models in structured reasoning use cases.",
+ "description": "Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It is optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories. The model features 480 billion total parameters, with 35 billion active per forward pass (8 out of 160 experts).\n\nPricing for the Alibaba endpoints varies by context length. Once a request is greater than 128k input tokens, the higher pricing is used.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27066,7 +27587,7 @@
"text"
],
"tokenizer": "Qwen3",
- "instruct_type": "qwen3"
+ "instruct_type": null
},
"top_provider": {
"context_length": 262144,
@@ -27076,17 +27597,16 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -27097,12 +27617,12 @@
}
},
{
- "id": "qwen/qwen3-235b-a22b:free",
- "name": "Qwen: Qwen3 235B A22B (free)",
+ "id": "qwen/qwen3-coder:free",
+ "name": "Qwen: Qwen3 Coder (free)",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-04-28 23:29:17 +0200",
- "context_window": 131072,
+ "created_at": "2025-07-23 10:29:06 +1000",
+ "context_window": 262144,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -27116,12 +27636,11 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
"predicted_outputs"
],
"pricing": {},
"metadata": {
- "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.",
+ "description": "Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It is optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories. The model features 480 billion total parameters, with 35 billion active per forward pass (8 out of 160 experts).\n\nPricing for the Alibaba endpoints varies by context length. Once a request is greater than 128k input tokens, the higher pricing is used.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27131,28 +27650,24 @@
"text"
],
"tokenizer": "Qwen3",
- "instruct_type": "qwen3"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 262144,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -27163,12 +27678,12 @@
}
},
{
- "id": "qwen/qwen3-30b-a3b",
- "name": "Qwen: Qwen3 30B A3B",
+ "id": "qwen/qwq-32b",
+ "name": "Qwen: QwQ 32B",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-04-29 00:16:44 +0200",
- "context_window": 40960,
+ "created_at": "2025-03-06 08:06:54 +1100",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -27181,20 +27696,19 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01999188,
- "output_per_million": 0.0800064
+ "input_per_million": 0.075,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.",
+ "description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27203,11 +27717,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen3",
- "instruct_type": "qwen3"
+ "tokenizer": "Qwen",
+ "instruct_type": "qwq"
},
"top_provider": {
- "context_length": 40960,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -27227,8 +27741,6 @@
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -27236,13 +27748,13 @@
}
},
{
- "id": "qwen/qwen3-30b-a3b-instruct-2507",
- "name": "Qwen: Qwen3 30B A3B Instruct 2507",
+ "id": "qwen/qwq-32b-preview",
+ "name": "Qwen: QwQ 32B Preview",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-07-29 18:36:05 +0200",
- "context_window": 131072,
- "max_output_tokens": 32768,
+ "created_at": "2024-11-28 11:42:21 +1100",
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -27254,18 +27766,18 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.19999999999999998,
- "output_per_million": 0.7999999999999999
+ "output_per_million": 0.19999999999999998
}
}
},
"metadata": {
- "description": "Qwen3-30B-A3B-Instruct-2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference. It operates in non-thinking mode and is designed for high-quality instruction following, multilingual understanding, and agentic tool use. Post-trained on instruction data, it demonstrates competitive performance across reasoning (AIME, ZebraLogic), coding (MultiPL-E, LiveCodeBench), and alignment (IFEval, WritingBench) benchmarks. It outperforms its non-instruct variant on subjective and open-ended tasks while retaining strong factual and coding performance.",
+ "description": "QwQ-32B-Preview is an experimental research model focused on AI reasoning capabilities developed by the Qwen Team. As a preview release, it demonstrates promising analytical abilities while having several important limitations:\n\n1. **Language Mixing and Code-Switching**: The model may mix languages or switch between them unexpectedly, affecting response clarity.\n2. **Recursive Reasoning Loops**: The model may enter circular reasoning patterns, leading to lengthy responses without a conclusive answer.\n3. **Safety and Ethical Considerations**: The model requires enhanced safety measures to ensure reliable and secure performance, and users should exercise caution when deploying it.\n4. **Performance and Benchmark Limitations**: The model excels in math and coding but has room for improvement in other areas, such as common sense reasoning and nuanced language understanding.\n\n",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27274,32 +27786,39 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen3",
- "instruct_type": null
+ "tokenizer": "Qwen",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 32768,
+ "context_length": 32768,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "repetition_penalty",
"seed",
+ "stop",
"temperature",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen3-30b-a3b:free",
- "name": "Qwen: Qwen3 30B A3B (free)",
+ "id": "qwen/qwq-32b:free",
+ "name": "Qwen: QwQ 32B (free)",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-04-29 00:16:44 +0200",
- "context_window": 40960,
+ "created_at": "2025-03-06 08:06:54 +1100",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -27312,11 +27831,11 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {},
"metadata": {
- "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.",
+ "description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27325,41 +27844,35 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen3",
- "instruct_type": "qwen3"
+ "tokenizer": "Qwen",
+ "instruct_type": "qwq"
},
"top_provider": {
- "context_length": 40960,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "reasoning",
- "repetition_penalty",
- "seed",
+ "response_format",
"stop",
+ "structured_outputs",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen3-32b",
- "name": "Qwen: Qwen3 32B",
+ "id": "raifle/sorcererlm-8x22b",
+ "name": "SorcererLM 8x22B",
"provider": "openrouter",
- "family": "qwen",
- "created_at": "2025-04-28 23:32:25 +0200",
- "context_window": 40960,
+ "family": "raifle",
+ "created_at": "2024-11-09 09:31:23 +1100",
+ "context_window": 16000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -27372,20 +27885,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.017992691999999998,
- "output_per_million": 0.07200576
+ "input_per_million": 4.5,
+ "output_per_million": 4.5
}
}
},
"metadata": {
- "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ",
+ "description": "SorcererLM is an advanced RP and storytelling model, built as a Low-rank 16-bit LoRA fine-tuned on [WizardLM-2 8x22B](/microsoft/wizardlm-2-8x22b).\n\n- Advanced reasoning and emotional intelligence for engaging and immersive interactions\n- Vivid writing capabilities enriched with spatial and contextual awareness\n- Enhanced narrative depth, promoting creative and dynamic storytelling",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27394,45 +27905,37 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen3",
- "instruct_type": "qwen3"
+ "tokenizer": "Mistral",
+ "instruct_type": "vicuna"
},
"top_provider": {
- "context_length": 40960,
+ "context_length": 16000,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen3-4b:free",
- "name": "Qwen: Qwen3 4B (free)",
+ "id": "rekaai/reka-flash-3:free",
+ "name": "Reka: Flash 3 (free)",
"provider": "openrouter",
- "family": "qwen",
- "created_at": "2025-04-30 18:38:24 +0200",
- "context_window": 40960,
+ "family": "rekaai",
+ "created_at": "2025-03-13 07:53:33 +1100",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -27445,12 +27948,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {},
"metadata": {
- "description": "Qwen3-4B is a 4 billion parameter dense language model from the Qwen3 series, designed to support both general-purpose and reasoning-intensive tasks. It introduces a dual-mode architecture—thinking and non-thinking—allowing dynamic switching between high-precision logical reasoning and efficient dialogue generation. This makes it well-suited for multi-turn chat, instruction following, and complex agent workflows.",
+ "description": "Reka Flash 3 is a general-purpose, instruction-tuned large language model with 21 billion parameters, developed by Reka. It excels at general chat, coding tasks, instruction-following, and function calling. Featuring a 32K context length and optimized through reinforcement learning (RLOO), it provides competitive performance comparable to proprietary models within a smaller parameter footprint. Ideal for low-latency, local, or on-device deployments, Reka Flash 3 is compact, supports efficient quantization (down to 11GB at 4-bit precision), and employs explicit reasoning tags (\"\") to indicate its internal thought process.\n\nReka Flash 3 is primarily an English model with limited multilingual understanding capabilities. The model weights are released under the Apache 2.0 license.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27459,11 +27961,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen3",
- "instruct_type": "qwen3"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 40960,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -27471,28 +27973,30 @@
"supported_parameters": [
"frequency_penalty",
"include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
"reasoning",
- "response_format",
+ "repetition_penalty",
+ "seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen3-8b",
- "name": "Qwen: Qwen3 8B",
+ "id": "sao10k/l3-euryale-70b",
+ "name": "Sao10k: Llama 3 Euryale 70B v2.1",
"provider": "openrouter",
- "family": "qwen",
- "created_at": "2025-04-28 23:43:52 +0200",
- "context_window": 128000,
- "max_output_tokens": 20000,
+ "family": "sao10k",
+ "created_at": "2024-06-18 10:00:00 +1000",
+ "context_window": 8192,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -27509,13 +28013,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.035,
- "output_per_million": 0.13799999999999998
+ "input_per_million": 1.48,
+ "output_per_million": 1.48
}
}
},
"metadata": {
- "description": "Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between \"thinking\" mode for math, coding, and logical inference, and \"non-thinking\" mode for general conversation. The model is fine-tuned for instruction-following, agent integration, creative writing, and multilingual use across 100+ languages and dialects. It natively supports a 32K token context window and can extend to 131K tokens with YaRN scaling.",
+ "description": "Euryale 70B v2.1 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k).\n\n- Better prompt adherence.\n- Better anatomy / spatial awareness.\n- Adapts much better to unique and custom formatting / reply formats.\n- Very creative, lots of unique swipes.\n- Is not restrictive during roleplays.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27524,23 +28028,21 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen3",
- "instruct_type": "qwen3"
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 20000,
+ "context_length": 8192,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
"seed",
"stop",
@@ -27551,13 +28053,13 @@
}
},
{
- "id": "qwen/qwen3-8b:free",
- "name": "Qwen: Qwen3 8B (free)",
+ "id": "sao10k/l3-lunaris-8b",
+ "name": "Sao10K: Llama 3 8B Lunaris",
"provider": "openrouter",
- "family": "qwen",
- "created_at": "2025-04-28 23:43:52 +0200",
- "context_window": 40960,
- "max_output_tokens": 40960,
+ "family": "sao10k",
+ "created_at": "2024-08-13 10:00:00 +1000",
+ "context_window": 8192,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -27569,11 +28071,19 @@
},
"capabilities": [
"streaming",
+ "structured_output",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.02,
+ "output_per_million": 0.049999999999999996
+ }
+ }
+ },
"metadata": {
- "description": "Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between \"thinking\" mode for math, coding, and logical inference, and \"non-thinking\" mode for general conversation. The model is fine-tuned for instruction-following, agent integration, creative writing, and multilingual use across 100+ languages and dialects. It natively supports a 32K token context window and can extend to 131K tokens with YaRN scaling.",
+ "description": "Lunaris 8B is a versatile generalist and roleplaying model based on Llama 3. It's a strategic merge of multiple models, designed to balance creativity with improved logic and general knowledge.\n\nCreated by [Sao10k](https://huggingface.co/Sao10k), this model aims to offer an improved experience over Stheno v3.2, with enhanced creativity and logical reasoning.\n\nFor best results, use with Llama 3 Instruct context template, temperature 1.4, and min_p 0.1.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27582,41 +28092,38 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen3",
- "instruct_type": "qwen3"
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 40960,
- "max_completion_tokens": 40960,
+ "context_length": 8192,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen3-coder",
- "name": "Qwen: Qwen3 Coder ",
+ "id": "sao10k/l3.1-euryale-70b",
+ "name": "Sao10K: Llama 3.1 Euryale 70B v2.2",
"provider": "openrouter",
- "family": "qwen",
- "created_at": "2025-07-23 02:29:06 +0200",
- "context_window": 262144,
+ "family": "sao10k",
+ "created_at": "2024-08-28 10:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -27629,20 +28136,19 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.7999999999999999
+ "input_per_million": 0.65,
+ "output_per_million": 0.75
}
}
},
"metadata": {
- "description": "Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It is optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories. The model features 480 billion total parameters, with 35 billion active per forward pass (8 out of 160 experts).\n\nPricing for the Alibaba endpoints varies by context length. Once a request is greater than 128k input tokens, the higher pricing is used.",
+ "description": "Euryale L3.1 70B v2.2 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.1](/models/sao10k/l3-euryale-70b).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27651,11 +28157,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen3",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 262144,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -27663,7 +28169,6 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -27673,22 +28178,19 @@
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwen3-coder:free",
- "name": "Qwen: Qwen3 Coder (free)",
+ "id": "sao10k/l3.3-euryale-70b",
+ "name": "Sao10K: Llama 3.3 Euryale 70B",
"provider": "openrouter",
- "family": "qwen",
- "created_at": "2025-07-23 02:29:06 +0200",
- "context_window": 262144,
- "max_output_tokens": null,
+ "family": "sao10k",
+ "created_at": "2024-12-19 02:32:08 +1100",
+ "context_window": 131072,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -27700,12 +28202,19 @@
},
"capabilities": [
"streaming",
- "function_calling",
+ "structured_output",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.65,
+ "output_per_million": 0.75
+ }
+ }
+ },
"metadata": {
- "description": "Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It is optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories. The model features 480 billion total parameters, with 35 billion active per forward pass (8 out of 160 experts).\n\nPricing for the Alibaba endpoints varies by context length. Once a request is greater than 128k input tokens, the higher pricing is used.",
+ "description": "Euryale L3.3 70B is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.2](/models/sao10k/l3-euryale-70b).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27714,41 +28223,39 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen3",
- "instruct_type": null
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 262144,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwq-32b",
- "name": "Qwen: QwQ 32B",
+ "id": "sarvamai/sarvam-m:free",
+ "name": "Sarvam AI: Sarvam-M (free)",
"provider": "openrouter",
- "family": "qwen",
- "created_at": "2025-03-05 22:06:54 +0100",
- "context_window": 131072,
+ "family": "sarvamai",
+ "created_at": "2025-05-26 01:53:33 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -27761,19 +28268,11 @@
},
"capabilities": [
"streaming",
- "structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.",
+ "description": "Sarvam-M is a 24 B-parameter, instruction-tuned derivative of Mistral-Small-3.1-24B-Base-2503, post-trained on English plus eleven major Indic languages (bn, hi, kn, gu, mr, ml, or, pa, ta, te). The model introduces a dual-mode interface: “non-think” for low-latency chat and a optional “think” phase that exposes chain-of-thought tokens for more demanding reasoning, math, and coding tasks. \n\nBenchmark reports show solid gains versus similarly sized open models on Indic-language QA, GSM-8K math, and SWE-Bench coding, making Sarvam-M a practical general-purpose choice for multilingual conversational agents as well as analytical workloads that mix English, native Indic scripts, or romanized text.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27782,29 +28281,25 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "qwq"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"top_k",
"top_logprobs",
@@ -27813,12 +28308,12 @@
}
},
{
- "id": "qwen/qwq-32b-preview",
- "name": "Qwen: QwQ 32B Preview",
+ "id": "scb10x/llama3.1-typhoon2-70b-instruct",
+ "name": "Typhoon2 70B Instruct",
"provider": "openrouter",
- "family": "qwen",
- "created_at": "2024-11-28 01:42:21 +0100",
- "context_window": 32768,
+ "family": "scb10x",
+ "created_at": "2025-03-29 08:09:30 +1100",
+ "context_window": 8192,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -27836,13 +28331,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.19999999999999998
+ "input_per_million": 0.88,
+ "output_per_million": 0.88
}
}
},
"metadata": {
- "description": "QwQ-32B-Preview is an experimental research model focused on AI reasoning capabilities developed by the Qwen Team. As a preview release, it demonstrates promising analytical abilities while having several important limitations:\n\n1. **Language Mixing and Code-Switching**: The model may mix languages or switch between them unexpectedly, affecting response clarity.\n2. **Recursive Reasoning Loops**: The model may enter circular reasoning patterns, leading to lengthy responses without a conclusive answer.\n3. **Safety and Ethical Considerations**: The model requires enhanced safety measures to ensure reliable and secure performance, and users should exercise caution when deploying it.\n4. **Performance and Benchmark Limitations**: The model excels in math and coding but has room for improvement in other areas, such as common sense reasoning and nuanced language understanding.\n\n",
+ "description": "Llama3.1-Typhoon2-70B-Instruct is a Thai-English instruction-tuned language model with 70 billion parameters, built on Llama 3.1. It demonstrates strong performance across general instruction-following, math, coding, and tool-use tasks, with state-of-the-art results in Thai-specific benchmarks such as IFEval, MT-Bench, and Thai-English code-switching.\n\nThe model excels in bilingual reasoning and function-calling scenarios, offering high accuracy across diverse domains. Comparative evaluations show consistent improvements over prior Thai LLMs and other Llama-based baselines. Full results and methodology are available in the [technical report.](https://arxiv.org/abs/2412.13702)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27851,11 +28346,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Llama3",
+ "instruct_type": "llama3"
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 8192,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -27863,26 +28358,23 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
- "seed",
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwq-32b:free",
- "name": "Qwen: QwQ 32B (free)",
+ "id": "shisa-ai/shisa-v2-llama3.3-70b",
+ "name": "Shisa AI: Shisa V2 Llama 3.3 70B ",
"provider": "openrouter",
- "family": "qwen",
- "created_at": "2025-03-05 22:06:54 +0100",
+ "family": "shisa-ai",
+ "created_at": "2025-04-16 08:07:38 +1000",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -27896,11 +28388,18 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.01999188,
+ "output_per_million": 0.0800064
+ }
+ }
+ },
"metadata": {
- "description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.",
+ "description": "Shisa V2 Llama 3.3 70B is a bilingual Japanese-English chat model fine-tuned by Shisa.AI on Meta’s Llama-3.3-70B-Instruct base. It prioritizes Japanese language performance while retaining strong English capabilities. The model was optimized entirely through post-training, using a refined mix of supervised fine-tuning (SFT) and DPO datasets including regenerated ShareGPT-style data, translation tasks, roleplaying conversations, and instruction-following prompts. Unlike earlier Shisa releases, this version avoids tokenizer modifications or extended pretraining.\n\nShisa V2 70B achieves leading Japanese task performance across a wide range of custom and public benchmarks, including JA MT Bench, ELYZA 100, and Rakuda. It supports a 128K token context length and integrates smoothly with inference frameworks like vLLM and SGLang. While it inherits safety characteristics from its base model, no additional alignment was applied. The model is intended for high-performance bilingual chat, instruction following, and translation tasks across JA/EN.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27909,8 +28408,8 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "qwq"
+ "tokenizer": "Llama3",
+ "instruct_type": null
},
"top_provider": {
"context_length": 32768,
@@ -27920,24 +28419,28 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "repetition_penalty",
+ "seed",
"stop",
- "structured_outputs",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "raifle/sorcererlm-8x22b",
- "name": "SorcererLM 8x22B",
+ "id": "shisa-ai/shisa-v2-llama3.3-70b:free",
+ "name": "Shisa AI: Shisa V2 Llama 3.3 70B (free)",
"provider": "openrouter",
- "family": "raifle",
- "created_at": "2024-11-08 23:31:23 +0100",
- "context_window": 16000,
+ "family": "shisa-ai",
+ "created_at": "2025-04-16 08:07:38 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -27948,20 +28451,13 @@
"text"
]
},
- "capabilities": [
- "streaming",
- "predicted_outputs"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 4.5,
- "output_per_million": 4.5
- }
- }
- },
+ "capabilities": [
+ "streaming",
+ "predicted_outputs"
+ ],
+ "pricing": {},
"metadata": {
- "description": "SorcererLM is an advanced RP and storytelling model, built as a Low-rank 16-bit LoRA fine-tuned on [WizardLM-2 8x22B](/microsoft/wizardlm-2-8x22b).\n\n- Advanced reasoning and emotional intelligence for engaging and immersive interactions\n- Vivid writing capabilities enriched with spatial and contextual awareness\n- Enhanced narrative depth, promoting creative and dynamic storytelling",
+ "description": "Shisa V2 Llama 3.3 70B is a bilingual Japanese-English chat model fine-tuned by Shisa.AI on Meta’s Llama-3.3-70B-Instruct base. It prioritizes Japanese language performance while retaining strong English capabilities. The model was optimized entirely through post-training, using a refined mix of supervised fine-tuning (SFT) and DPO datasets including regenerated ShareGPT-style data, translation tasks, roleplaying conversations, and instruction-following prompts. Unlike earlier Shisa releases, this version avoids tokenizer modifications or extended pretraining.\n\nShisa V2 70B achieves leading Japanese task performance across a wide range of custom and public benchmarks, including JA MT Bench, ELYZA 100, and Rakuda. It supports a 128K token context length and integrates smoothly with inference frameworks like vLLM and SGLang. While it inherits safety characteristics from its base model, no additional alignment was applied. The model is intended for high-performance bilingual chat, instruction following, and translation tasks across JA/EN.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -27970,11 +28466,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
- "instruct_type": "vicuna"
+ "tokenizer": "Llama3",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 16000,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -27982,6 +28478,7 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -27990,18 +28487,19 @@
"stop",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "rekaai/reka-flash-3:free",
- "name": "Reka: Flash 3 (free)",
+ "id": "sophosympatheia/midnight-rose-70b",
+ "name": "Midnight Rose 70B",
"provider": "openrouter",
- "family": "rekaai",
- "created_at": "2025-03-12 21:53:33 +0100",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "sophosympatheia",
+ "created_at": "2024-03-22 11:00:00 +1100",
+ "context_window": 4096,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -28015,9 +28513,16 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.7999999999999999,
+ "output_per_million": 0.7999999999999999
+ }
+ }
+ },
"metadata": {
- "description": "Reka Flash 3 is a general-purpose, instruction-tuned large language model with 21 billion parameters, developed by Reka. It excels at general chat, coding tasks, instruction-following, and function calling. Featuring a 32K context length and optimized through reinforcement learning (RLOO), it provides competitive performance comparable to proprietary models within a smaller parameter footprint. Ideal for low-latency, local, or on-device deployments, Reka Flash 3 is compact, supports efficient quantization (down to 11GB at 4-bit precision), and employs explicit reasoning tags (\"\") to indicate its internal thought process.\n\nReka Flash 3 is primarily an English model with limited multilingual understanding capabilities. The model weights are released under the Apache 2.0 license.",
+ "description": "A merge with a complex family tree, this model was crafted for roleplaying and storytelling. Midnight Rose is a successor to Rogue Rose and Aurora Nights and improves upon them both. It wants to produce lengthy output by default and is the best creative writing merge produced so far by sophosympatheia.\n\nDescending from earlier versions of Midnight Rose and [Wizard Tulu Dolphin 70B](https://huggingface.co/sophosympatheia/Wizard-Tulu-Dolphin-70B-v1.0), it inherits the best qualities of each.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28026,42 +28531,38 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": null
+ "tokenizer": "Llama2",
+ "instruct_type": "airoboros"
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 4096,
+ "max_completion_tokens": 2048,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
"seed",
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "sao10k/l3-euryale-70b",
- "name": "Sao10k: Llama 3 Euryale 70B v2.1",
+ "id": "switchpoint/router",
+ "name": "Switchpoint Router",
"provider": "openrouter",
- "family": "sao10k",
- "created_at": "2024-06-18 02:00:00 +0200",
- "context_window": 8192,
- "max_output_tokens": 8192,
+ "family": "switchpoint",
+ "created_at": "2025-07-12 08:28:19 +1000",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -28072,19 +28573,18 @@
]
},
"capabilities": [
- "streaming",
- "predicted_outputs"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.48,
- "output_per_million": 1.48
+ "input_per_million": 0.85,
+ "output_per_million": 3.4
}
}
},
"metadata": {
- "description": "Euryale 70B v2.1 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k).\n\n- Better prompt adherence.\n- Better anatomy / spatial awareness.\n- Adapts much better to unique and custom formatting / reply formats.\n- Very creative, lots of unique swipes.\n- Is not restrictive during roleplays.",
+ "description": "Switchpoint AI's router instantly analyzes your request and directs it to the optimal AI from an ever-evolving library. \n\nAs the world of LLMs advances, our router gets smarter, ensuring you always benefit from the industry's newest models without changing your workflow.\n\nThis model is configured for a simple, flat rate per response here on OpenRouter. It's powered by the full routing engine from [Switchpoint AI](https://www.switchpoint.dev).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28093,22 +28593,19 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 8192,
+ "context_length": 131072,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
+ "include_reasoning",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
+ "reasoning",
"seed",
"stop",
"temperature",
@@ -28118,12 +28615,12 @@
}
},
{
- "id": "sao10k/l3-lunaris-8b",
- "name": "Sao10K: Llama 3 8B Lunaris",
+ "id": "tencent/hunyuan-a13b-instruct",
+ "name": "Tencent: Hunyuan A13B Instruct",
"provider": "openrouter",
- "family": "sao10k",
- "created_at": "2024-08-13 02:00:00 +0200",
- "context_window": 8192,
+ "family": "tencent",
+ "created_at": "2025-07-09 01:14:24 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -28136,19 +28633,18 @@
},
"capabilities": [
"streaming",
- "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.02,
- "output_per_million": 0.049999999999999996
+ "input_per_million": 0.03,
+ "output_per_million": 0.03
}
}
},
"metadata": {
- "description": "Lunaris 8B is a versatile generalist and roleplaying model based on Llama 3. It's a strategic merge of multiple models, designed to balance creativity with improved logic and general knowledge.\n\nCreated by [Sao10k](https://huggingface.co/Sao10k), this model aims to offer an improved experience over Stheno v3.2, with enhanced creativity and logical reasoning.\n\nFor best results, use with Llama 3 Instruct context template, temperature 1.4, and min_p 0.1.",
+ "description": "Hunyuan-A13B is a 13B active parameter Mixture-of-Experts (MoE) language model developed by Tencent, with a total parameter count of 80B and support for reasoning via Chain-of-Thought. It offers competitive benchmark performance across mathematics, science, coding, and multi-turn reasoning tasks, while maintaining high inference efficiency via Grouped Query Attention (GQA) and quantization support (FP8, GPTQ, etc.).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28157,37 +28653,40 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 8192,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "sao10k/l3.1-euryale-70b",
- "name": "Sao10K: Llama 3.1 Euryale 70B v2.2",
+ "id": "tencent/hunyuan-a13b-instruct:free",
+ "name": "Tencent: Hunyuan A13B Instruct (free)",
"provider": "openrouter",
- "family": "sao10k",
- "created_at": "2024-08-28 02:00:00 +0200",
+ "family": "tencent",
+ "created_at": "2025-07-09 01:14:24 +1000",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -28201,19 +28700,11 @@
},
"capabilities": [
"streaming",
- "structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.65,
- "output_per_million": 0.75
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Euryale L3.1 70B v2.2 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.1](/models/sao10k/l3-euryale-70b).",
+ "description": "Hunyuan-A13B is a 13B active parameter Mixture-of-Experts (MoE) language model developed by Tencent, with a total parameter count of 80B and support for reasoning via Chain-of-Thought. It offers competitive benchmark performance across mathematics, science, coding, and multi-turn reasoning tasks, while maintaining high inference efficiency via Grouped Query Attention (GQA) and quantization support (FP8, GPTQ, etc.).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28222,8 +28713,8 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
"context_length": 32768,
@@ -28233,29 +28724,31 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "sao10k/l3.3-euryale-70b",
- "name": "Sao10K: Llama 3.3 Euryale 70B",
+ "id": "thedrummer/anubis-70b-v1.1",
+ "name": "TheDrummer: Anubis 70B V1.1",
"provider": "openrouter",
- "family": "sao10k",
- "created_at": "2024-12-18 16:32:08 +0100",
- "context_window": 131072,
- "max_output_tokens": 16384,
+ "family": "thedrummer",
+ "created_at": "2025-06-30 00:45:47 +1000",
+ "context_window": 16384,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -28267,19 +28760,18 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.65,
- "output_per_million": 0.75
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 0.7
}
}
},
"metadata": {
- "description": "Euryale L3.3 70B is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.2](/models/sao10k/l3-euryale-70b).",
+ "description": "TheDrummer's Anubis v1.1 is an unaligned, creative Llama 3.3 70B model focused on providing character-driven roleplay & stories. It excels at gritty, visceral prose, unique character adherence, and coherent narratives, while maintaining the instruction following Llama 3.3 70B is known for.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28289,17 +28781,16 @@
"text"
],
"tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 16384,
+ "context_length": 16384,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"max_tokens",
"min_p",
"presence_penalty",
@@ -28315,13 +28806,13 @@
}
},
{
- "id": "sarvamai/sarvam-m:free",
- "name": "Sarvam AI: Sarvam-M (free)",
+ "id": "thedrummer/anubis-pro-105b-v1",
+ "name": "TheDrummer: Anubis Pro 105B V1",
"provider": "openrouter",
- "family": "sarvamai",
- "created_at": "2025-05-25 17:53:33 +0200",
- "context_window": 32768,
- "max_output_tokens": null,
+ "family": "thedrummer",
+ "created_at": "2025-03-11 08:31:30 +1100",
+ "context_window": 131072,
+ "max_output_tokens": 131072,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -28332,12 +28823,18 @@
]
},
"capabilities": [
- "streaming",
- "predicted_outputs"
+ "streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.5,
+ "output_per_million": 1.0
+ }
+ }
+ },
"metadata": {
- "description": "Sarvam-M is a 24 B-parameter, instruction-tuned derivative of Mistral-Small-3.1-24B-Base-2503, post-trained on English plus eleven major Indic languages (bn, hi, kn, gu, mr, ml, or, pa, ta, te). The model introduces a dual-mode interface: “non-think” for low-latency chat and a optional “think” phase that exposes chain-of-thought tokens for more demanding reasoning, math, and coding tasks. \n\nBenchmark reports show solid gains versus similarly sized open models on Indic-language QA, GSM-8K math, and SWE-Bench coding, making Sarvam-M a practical general-purpose choice for multilingual conversational agents as well as analytical workloads that mix English, native Indic scripts, or romanized text.",
+ "description": "Anubis Pro 105B v1 is an expanded and refined variant of Meta’s Llama 3.3 70B, featuring 50% additional layers and further fine-tuning to leverage its increased capacity. Designed for advanced narrative, roleplay, and instructional tasks, it demonstrates enhanced emotional intelligence, creativity, nuanced character portrayal, and superior prompt adherence compared to smaller models. Its larger parameter count allows for deeper contextual understanding and extended reasoning capabilities, optimized for engaging, intelligent, and coherent interactions.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28350,15 +28847,13 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 131072,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -28367,18 +28862,17 @@
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "scb10x/llama3.1-typhoon2-70b-instruct",
- "name": "Typhoon2 70B Instruct",
+ "id": "thedrummer/rocinante-12b",
+ "name": "TheDrummer: Rocinante 12B",
"provider": "openrouter",
- "family": "scb10x",
- "created_at": "2025-03-28 22:09:30 +0100",
- "context_window": 8192,
+ "family": "thedrummer",
+ "created_at": "2024-09-30 10:00:00 +1000",
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -28391,18 +28885,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.88,
- "output_per_million": 0.88
+ "input_per_million": 0.16999999999999998,
+ "output_per_million": 0.43
}
}
},
"metadata": {
- "description": "Llama3.1-Typhoon2-70B-Instruct is a Thai-English instruction-tuned language model with 70 billion parameters, built on Llama 3.1. It demonstrates strong performance across general instruction-following, math, coding, and tool-use tasks, with state-of-the-art results in Thai-specific benchmarks such as IFEval, MT-Bench, and Thai-English code-switching.\n\nThe model excels in bilingual reasoning and function-calling scenarios, offering high accuracy across diverse domains. Comparative evaluations show consistent improvements over prior Thai LLMs and other Llama-based baselines. Full results and methodology are available in the [technical report.](https://arxiv.org/abs/2412.13702)",
+ "description": "Rocinante 12B is designed for engaging storytelling and rich prose.\n\nEarly testers have reported:\n- Expanded vocabulary with unique and expressive word choices\n- Enhanced creativity for vivid narratives\n- Adventure-filled and captivating stories",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28411,11 +28907,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
+ "tokenizer": "Qwen",
+ "instruct_type": "chatml"
},
"top_provider": {
- "context_length": 8192,
+ "context_length": 32768,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -28423,23 +28919,29 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
+ "seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "shisa-ai/shisa-v2-llama3.3-70b",
- "name": "Shisa AI: Shisa V2 Llama 3.3 70B ",
+ "id": "thedrummer/skyfall-36b-v2",
+ "name": "TheDrummer: Skyfall 36B V2",
"provider": "openrouter",
- "family": "shisa-ai",
- "created_at": "2025-04-16 00:07:38 +0200",
+ "family": "thedrummer",
+ "created_at": "2025-03-11 06:56:06 +1100",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -28453,18 +28955,19 @@
},
"capabilities": [
"streaming",
+ "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01999188,
- "output_per_million": 0.0800064
+ "input_per_million": 0.0481286,
+ "output_per_million": 0.192608
}
}
},
"metadata": {
- "description": "Shisa V2 Llama 3.3 70B is a bilingual Japanese-English chat model fine-tuned by Shisa.AI on Meta’s Llama-3.3-70B-Instruct base. It prioritizes Japanese language performance while retaining strong English capabilities. The model was optimized entirely through post-training, using a refined mix of supervised fine-tuning (SFT) and DPO datasets including regenerated ShareGPT-style data, translation tasks, roleplaying conversations, and instruction-following prompts. Unlike earlier Shisa releases, this version avoids tokenizer modifications or extended pretraining.\n\nShisa V2 70B achieves leading Japanese task performance across a wide range of custom and public benchmarks, including JA MT Bench, ELYZA 100, and Rakuda. It supports a 128K token context length and integrates smoothly with inference frameworks like vLLM and SGLang. While it inherits safety characteristics from its base model, no additional alignment was applied. The model is intended for high-performance bilingual chat, instruction following, and translation tasks across JA/EN.",
+ "description": "Skyfall 36B v2 is an enhanced iteration of Mistral Small 2501, specifically fine-tuned for improved creativity, nuanced writing, role-playing, and coherent storytelling.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28473,7 +28976,7 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
@@ -28490,8 +28993,10 @@
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"top_k",
"top_logprobs",
@@ -28500,11 +29005,11 @@
}
},
{
- "id": "shisa-ai/shisa-v2-llama3.3-70b:free",
- "name": "Shisa AI: Shisa V2 Llama 3.3 70B (free)",
+ "id": "thedrummer/unslopnemo-12b",
+ "name": "TheDrummer: UnslopNemo 12B",
"provider": "openrouter",
- "family": "shisa-ai",
- "created_at": "2025-04-16 00:07:38 +0200",
+ "family": "thedrummer",
+ "created_at": "2024-11-09 09:04:08 +1100",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -28518,11 +29023,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 0.39999999999999997
+ }
+ }
+ },
"metadata": {
- "description": "Shisa V2 Llama 3.3 70B is a bilingual Japanese-English chat model fine-tuned by Shisa.AI on Meta’s Llama-3.3-70B-Instruct base. It prioritizes Japanese language performance while retaining strong English capabilities. The model was optimized entirely through post-training, using a refined mix of supervised fine-tuning (SFT) and DPO datasets including regenerated ShareGPT-style data, translation tasks, roleplaying conversations, and instruction-following prompts. Unlike earlier Shisa releases, this version avoids tokenizer modifications or extended pretraining.\n\nShisa V2 70B achieves leading Japanese task performance across a wide range of custom and public benchmarks, including JA MT Bench, ELYZA 100, and Rakuda. It supports a 128K token context length and integrates smoothly with inference frameworks like vLLM and SGLang. While it inherits safety characteristics from its base model, no additional alignment was applied. The model is intended for high-performance bilingual chat, instruction following, and translation tasks across JA/EN.",
+ "description": "UnslopNemo v4.1 is the latest addition from the creator of Rocinante, designed for adventure writing and role-play scenarios.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28531,8 +29045,8 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": null
+ "tokenizer": "Mistral",
+ "instruct_type": "mistral"
},
"top_provider": {
"context_length": 32768,
@@ -28548,23 +29062,26 @@
"min_p",
"presence_penalty",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "sophosympatheia/midnight-rose-70b",
- "name": "Midnight Rose 70B",
+ "id": "thudm/glm-4-32b",
+ "name": "THUDM: GLM 4 32B",
"provider": "openrouter",
- "family": "sophosympatheia",
- "created_at": "2024-03-22 01:00:00 +0100",
- "context_window": 4096,
- "max_output_tokens": 2048,
+ "family": "thudm",
+ "created_at": "2025-04-18 06:15:15 +1000",
+ "context_window": 32000,
+ "max_output_tokens": 32000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -28581,13 +29098,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7999999999999999,
- "output_per_million": 0.7999999999999999
+ "input_per_million": 0.55,
+ "output_per_million": 1.66
}
}
},
"metadata": {
- "description": "A merge with a complex family tree, this model was crafted for roleplaying and storytelling. Midnight Rose is a successor to Rogue Rose and Aurora Nights and improves upon them both. It wants to produce lengthy output by default and is the best creative writing merge produced so far by sophosympatheia.\n\nDescending from earlier versions of Midnight Rose and [Wizard Tulu Dolphin 70B](https://huggingface.co/sophosympatheia/Wizard-Tulu-Dolphin-70B-v1.0), it inherits the best qualities of each.",
+ "description": "GLM-4-32B-0414 is a 32B bilingual (Chinese-English) open-weight language model optimized for code generation, function calling, and agent-style tasks. Pretrained on 15T of high-quality and reasoning-heavy data, it was further refined using human preference alignment, rejection sampling, and reinforcement learning. The model excels in complex reasoning, artifact generation, and structured output tasks, achieving performance comparable to GPT-4o and DeepSeek-V3-0324 across several benchmarks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28596,12 +29113,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama2",
- "instruct_type": "airoboros"
+ "tokenizer": "Other",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 4096,
- "max_completion_tokens": 2048,
+ "context_length": 32000,
+ "max_completion_tokens": 32000,
"is_moderated": false
},
"per_request_limits": null,
@@ -28621,16 +29138,17 @@
}
},
{
- "id": "switchpoint/router",
- "name": "Switchpoint Router",
+ "id": "thudm/glm-4.1v-9b-thinking",
+ "name": "THUDM: GLM 4.1V 9B Thinking",
"provider": "openrouter",
- "family": "switchpoint",
- "created_at": "2025-07-12 00:28:19 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "family": "thudm",
+ "created_at": "2025-07-12 00:33:05 +1000",
+ "context_window": 65536,
+ "max_output_tokens": 8000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -28638,21 +29156,23 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.85,
- "output_per_million": 3.4
+ "input_per_million": 0.035,
+ "output_per_million": 0.13799999999999998
}
}
},
"metadata": {
- "description": "Switchpoint AI's router instantly analyzes your request and directs it to the optimal AI from an ever-evolving library. \n\nAs the world of LLMs advances, our router gets smarter, ensuring you always benefit from the industry's newest models without changing your workflow.\n\nThis model is configured for a simple, flat rate per response here on OpenRouter. It's powered by the full routing engine from [Switchpoint AI](https://www.switchpoint.dev).",
+ "description": "GLM-4.1V-9B-Thinking is a 9B parameter vision-language model developed by THUDM, based on the GLM-4-9B foundation. It introduces a reasoning-centric \"thinking paradigm\" enhanced with reinforcement learning to improve multimodal reasoning, long-context understanding (up to 64K tokens), and complex problem solving. It achieves state-of-the-art performance among models in its class, outperforming even larger models like Qwen-2.5-VL-72B on a majority of benchmark tasks. ",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
+ "image",
"text"
],
"output_modalities": [
@@ -28662,15 +29182,20 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
+ "context_length": 65536,
+ "max_completion_tokens": 8000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"include_reasoning",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
"reasoning",
+ "repetition_penalty",
"seed",
"stop",
"temperature",
@@ -28680,11 +29205,11 @@
}
},
{
- "id": "tencent/hunyuan-a13b-instruct",
- "name": "Tencent: Hunyuan A13B Instruct",
+ "id": "thudm/glm-z1-32b",
+ "name": "THUDM: GLM Z1 32B",
"provider": "openrouter",
- "family": "tencent",
- "created_at": "2025-07-08 17:14:24 +0200",
+ "family": "thudm",
+ "created_at": "2025-04-18 07:09:08 +1000",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -28703,13 +29228,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.03,
- "output_per_million": 0.03
+ "input_per_million": 0.01999188,
+ "output_per_million": 0.0800064
}
}
},
"metadata": {
- "description": "Hunyuan-A13B is a 13B active parameter Mixture-of-Experts (MoE) language model developed by Tencent, with a total parameter count of 80B and support for reasoning via Chain-of-Thought. It offers competitive benchmark performance across mathematics, science, coding, and multi-turn reasoning tasks, while maintaining high inference efficiency via Grouped Query Attention (GQA) and quantization support (FP8, GPTQ, etc.).",
+ "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28719,7 +29244,7 @@
"text"
],
"tokenizer": "Other",
- "instruct_type": null
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
"context_length": 32768,
@@ -28747,12 +29272,12 @@
}
},
{
- "id": "tencent/hunyuan-a13b-instruct:free",
- "name": "Tencent: Hunyuan A13B Instruct (free)",
+ "id": "tngtech/deepseek-r1t-chimera",
+ "name": "TNG: DeepSeek R1T Chimera",
"provider": "openrouter",
- "family": "tencent",
- "created_at": "2025-07-08 17:14:24 +0200",
- "context_window": 32768,
+ "family": "tngtech",
+ "created_at": "2025-04-27 23:34:35 +1000",
+ "context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -28767,9 +29292,16 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.1999188,
+ "output_per_million": 0.800064
+ }
+ }
+ },
"metadata": {
- "description": "Hunyuan-A13B is a 13B active parameter Mixture-of-Experts (MoE) language model developed by Tencent, with a total parameter count of 80B and support for reasoning via Chain-of-Thought. It offers competitive benchmark performance across mathematics, science, coding, and multi-turn reasoning tasks, while maintaining high inference efficiency via Grouped Query Attention (GQA) and quantization support (FP8, GPTQ, etc.).",
+ "description": "DeepSeek-R1T-Chimera is created by merging DeepSeek-R1 and DeepSeek-V3 (0324), combining the reasoning capabilities of R1 with the token efficiency improvements of V3. It is based on a DeepSeek-MoE Transformer architecture and is optimized for general text generation tasks.\n\nThe model merges pretrained weights from both source models to balance performance across reasoning, efficiency, and instruction-following tasks. It is released under the MIT license and intended for research and commercial use.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28778,11 +29310,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "DeepSeek",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 163840,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -28807,12 +29339,12 @@
}
},
{
- "id": "thedrummer/anubis-70b-v1.1",
- "name": "TheDrummer: Anubis 70B V1.1",
+ "id": "tngtech/deepseek-r1t-chimera:free",
+ "name": "TNG: DeepSeek R1T Chimera (free)",
"provider": "openrouter",
- "family": "thedrummer",
- "created_at": "2025-06-29 16:45:47 +0200",
- "context_window": 16384,
+ "family": "tngtech",
+ "created_at": "2025-04-27 23:34:35 +1000",
+ "context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -28825,18 +29357,11 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 0.7
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "TheDrummer's Anubis v1.1 is an unaligned, creative Llama 3.3 70B model focused on providing character-driven roleplay & stories. It excels at gritty, visceral prose, unique character adherence, and coherent narratives, while maintaining the instruction following Llama 3.3 70B is known for.",
+ "description": "DeepSeek-R1T-Chimera is created by merging DeepSeek-R1 and DeepSeek-V3 (0324), combining the reasoning capabilities of R1 with the token efficiency improvements of V3. It is based on a DeepSeek-MoE Transformer architecture and is optimized for general text generation tasks.\n\nThe model merges pretrained weights from both source models to balance performance across reasoning, efficiency, and instruction-following tasks. It is released under the MIT license and intended for research and commercial use.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28845,39 +29370,42 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
+ "tokenizer": "DeepSeek",
"instruct_type": null
},
"top_provider": {
- "context_length": 16384,
+ "context_length": 163840,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "thedrummer/anubis-pro-105b-v1",
- "name": "TheDrummer: Anubis Pro 105B V1",
+ "id": "tngtech/deepseek-r1t2-chimera:free",
+ "name": "TNG: DeepSeek R1T2 Chimera (free)",
"provider": "openrouter",
- "family": "thedrummer",
- "created_at": "2025-03-10 22:31:30 +0100",
- "context_window": 131072,
- "max_output_tokens": 131072,
+ "family": "tngtech",
+ "created_at": "2025-07-09 01:03:05 +1000",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -28888,18 +29416,12 @@
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.0
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Anubis Pro 105B v1 is an expanded and refined variant of Meta’s Llama 3.3 70B, featuring 50% additional layers and further fine-tuning to leverage its increased capacity. Designed for advanced narrative, roleplay, and instructional tasks, it demonstrates enhanced emotional intelligence, creativity, nuanced character portrayal, and superior prompt adherence compared to smaller models. Its larger parameter count allows for deeper contextual understanding and extended reasoning capabilities, optimized for engaging, intelligent, and coherent interactions.",
+ "description": "DeepSeek-TNG-R1T2-Chimera is the second-generation Chimera model from TNG Tech. It is a 671 B-parameter mixture-of-experts text-generation model assembled from DeepSeek-AI’s R1-0528, R1, and V3-0324 checkpoints with an Assembly-of-Experts merge. The tri-parent design yields strong reasoning performance while running roughly 20 % faster than the original R1 and more than 2× faster than R1-0528 under vLLM, giving a favorable cost-to-intelligence trade-off. The checkpoint supports contexts up to 60 k tokens in standard use (tested to ~130 k) and maintains consistent token behaviour, making it suitable for long-context analysis, dialogue and other open-ended generation tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28908,36 +29430,41 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "DeepSeek",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 131072,
+ "context_length": 163840,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
"seed",
"stop",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "thedrummer/rocinante-12b",
- "name": "TheDrummer: Rocinante 12B",
+ "id": "undi95/remm-slerp-l2-13b",
+ "name": "ReMM SLERP 13B",
"provider": "openrouter",
- "family": "thedrummer",
- "created_at": "2024-09-30 02:00:00 +0200",
- "context_window": 32768,
+ "family": "undi95",
+ "created_at": "2023-07-22 10:00:00 +1000",
+ "context_window": 6144,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -28950,20 +29477,19 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.16999999999999998,
- "output_per_million": 0.43
+ "input_per_million": 0.44999999999999996,
+ "output_per_million": 0.65
}
}
},
"metadata": {
- "description": "Rocinante 12B is designed for engaging storytelling and rich prose.\n\nEarly testers have reported:\n- Expanded vocabulary with unique and expressive word choices\n- Enhanced creativity for vivid narratives\n- Adventure-filled and captivating stories",
+ "description": "A recreation trial of the original MythoMax-L2-B13 but with updated models. #merge",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -28972,11 +29498,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "chatml"
+ "tokenizer": "Llama2",
+ "instruct_type": "alpaca"
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 6144,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -28984,7 +29510,6 @@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -28994,20 +29519,19 @@
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_a",
"top_k",
"top_p"
]
}
},
{
- "id": "thedrummer/skyfall-36b-v2",
- "name": "TheDrummer: Skyfall 36B V2",
+ "id": "x-ai/grok-2-1212",
+ "name": "xAI: Grok 2 1212",
"provider": "openrouter",
- "family": "thedrummer",
- "created_at": "2025-03-10 20:56:06 +0100",
- "context_window": 32768,
+ "family": "x-ai",
+ "created_at": "2024-12-15 14:20:14 +1100",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -29020,19 +29544,19 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.0481286,
- "output_per_million": 0.192608
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "Skyfall 36B v2 is an enhanced iteration of Mistral Small 2501, specifically fine-tuned for improved creativity, nuanced writing, role-playing, and coherent storytelling.",
+ "description": "Grok 2 1212 introduces significant enhancements to accuracy, instruction adherence, and multilingual support, making it a powerful and flexible choice for developers seeking a highly steerable, intelligent model.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -29041,46 +29565,44 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Grok",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "thedrummer/unslopnemo-12b",
- "name": "TheDrummer: UnslopNemo 12B",
+ "id": "x-ai/grok-2-vision-1212",
+ "name": "xAI: Grok 2 Vision 1212",
"provider": "openrouter",
- "family": "thedrummer",
- "created_at": "2024-11-08 23:04:08 +0100",
+ "family": "x-ai",
+ "created_at": "2024-12-15 15:35:38 +1100",
"context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -29088,30 +29610,29 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 0.39999999999999997
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "UnslopNemo v4.1 is the latest addition from the creator of Rocinante, designed for adventure writing and role-play scenarios.",
+ "description": "Grok 2 Vision 1212 advances image-based AI with stronger visual comprehension, refined instruction-following, and multilingual support. From object recognition to style analysis, it empowers developers to build more intuitive, visually aware applications. Its enhanced steerability and reasoning establish a robust foundation for next-generation image solutions.\n\nTo read more about this model, check out [xAI's announcement](https://x.ai/blog/grok-1212).",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Mistral",
- "instruct_type": "mistral"
+ "tokenizer": "Grok",
+ "instruct_type": null
},
"top_provider": {
"context_length": 32768,
@@ -29121,32 +29642,26 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
"logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
- "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "thudm/glm-4-32b",
- "name": "THUDM: GLM 4 32B",
+ "id": "x-ai/grok-3",
+ "name": "xAI: Grok 3",
"provider": "openrouter",
- "family": "thudm",
- "created_at": "2025-04-17 22:15:15 +0200",
- "context_window": 32000,
- "max_output_tokens": 32000,
+ "family": "x-ai",
+ "created_at": "2025-06-11 05:15:08 +1000",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -29158,18 +29673,20 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.55,
- "output_per_million": 1.66
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
}
}
},
"metadata": {
- "description": "GLM-4-32B-0414 is a 32B bilingual (Chinese-English) open-weight language model optimized for code generation, function calling, and agent-style tasks. Pretrained on 15T of high-quality and reasoning-heavy data, it was further refined using human preference alignment, rejection sampling, and reinforcement learning. The model excels in complex reasoning, artifact generation, and structured output tasks, achieving performance comparable to GPT-4o and DeepSeek-V3-0324 across several benchmarks.",
+ "description": "Grok 3 is the latest model from xAI. It's their flagship model that excels at enterprise use cases like data extraction, coding, and text summarization. Possesses deep domain knowledge in finance, healthcare, law, and science.\n\n",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -29178,42 +29695,43 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Grok",
"instruct_type": null
},
"top_provider": {
- "context_length": 32000,
- "max_completion_tokens": 32000,
+ "context_length": 131072,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
+ "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "thudm/glm-4.1v-9b-thinking",
- "name": "THUDM: GLM 4.1V 9B Thinking",
+ "id": "x-ai/grok-3-beta",
+ "name": "xAI: Grok 3 Beta",
"provider": "openrouter",
- "family": "thudm",
- "created_at": "2025-07-11 16:33:05 +0200",
- "context_window": 65536,
- "max_output_tokens": 8000,
+ "family": "x-ai",
+ "created_at": "2025-04-10 09:07:48 +1000",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -29222,60 +29740,60 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.035,
- "output_per_million": 0.13799999999999998
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
}
}
},
"metadata": {
- "description": "GLM-4.1V-9B-Thinking is a 9B parameter vision-language model developed by THUDM, based on the GLM-4-9B foundation. It introduces a reasoning-centric \"thinking paradigm\" enhanced with reinforcement learning to improve multimodal reasoning, long-context understanding (up to 64K tokens), and complex problem solving. It achieves state-of-the-art performance among models in its class, outperforming even larger models like Qwen-2.5-VL-72B on a majority of benchmark tasks. ",
+ "description": "Grok 3 is the latest model from xAI. It's their flagship model that excels at enterprise use cases like data extraction, coding, and text summarization. Possesses deep domain knowledge in finance, healthcare, law, and science.\n\nExcels in structured tasks and benchmarks like GPQA, LCB, and MMLU-Pro where it outperforms Grok 3 Mini even on high thinking. \n\nNote: That there are two xAI endpoints for this model. By default when using this model we will always route you to the base endpoint. If you want the fast endpoint you can add `provider: { sort: throughput}`, to sort by throughput instead. \n",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Grok",
"instruct_type": null
},
"top_provider": {
- "context_length": 65536,
- "max_completion_tokens": 8000,
+ "context_length": 131072,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
- "logit_bias",
+ "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "thudm/glm-z1-32b",
- "name": "THUDM: GLM Z1 32B",
+ "id": "x-ai/grok-3-mini",
+ "name": "xAI: Grok 3 Mini",
"provider": "openrouter",
- "family": "thudm",
- "created_at": "2025-04-17 23:09:08 +0200",
- "context_window": 32768,
+ "family": "x-ai",
+ "created_at": "2025-06-11 05:20:45 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -29288,18 +29806,20 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01999188,
- "output_per_million": 0.0800064
+ "input_per_million": 0.3,
+ "output_per_million": 0.5,
+ "cached_input_per_million": 0.075
}
}
},
"metadata": {
- "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.",
+ "description": "A lightweight model that thinks before responding. Fast, smart, and great for logic-based tasks that do not require deep domain knowledge. The raw thinking traces are accessible.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -29308,41 +29828,39 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Grok",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
"logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "tngtech/deepseek-r1t-chimera",
- "name": "TNG: DeepSeek R1T Chimera",
+ "id": "x-ai/grok-3-mini-beta",
+ "name": "xAI: Grok 3 Mini Beta",
"provider": "openrouter",
- "family": "tngtech",
- "created_at": "2025-04-27 15:34:35 +0200",
- "context_window": 163840,
+ "family": "x-ai",
+ "created_at": "2025-04-10 09:09:55 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -29355,18 +29873,20 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1999188,
- "output_per_million": 0.800064
+ "input_per_million": 0.3,
+ "output_per_million": 0.5,
+ "cached_input_per_million": 0.075
}
}
},
"metadata": {
- "description": "DeepSeek-R1T-Chimera is created by merging DeepSeek-R1 and DeepSeek-V3 (0324), combining the reasoning capabilities of R1 with the token efficiency improvements of V3. It is based on a DeepSeek-MoE Transformer architecture and is optimized for general text generation tasks.\n\nThe model merges pretrained weights from both source models to balance performance across reasoning, efficiency, and instruction-following tasks. It is released under the MIT license and intended for research and commercial use.",
+ "description": "Grok 3 Mini is a lightweight, smaller thinking model. Unlike traditional models that generate answers immediately, Grok 3 Mini thinks before responding. It’s ideal for reasoning-heavy tasks that don’t demand extensive domain knowledge, and shines in math-specific and quantitative use cases, such as solving challenging puzzles or math problems.\n\nTransparent \"thinking\" traces accessible. Defaults to low reasoning, can boost with setting `reasoning: { effort: \"high\" }`\n\nNote: That there are two xAI endpoints for this model. By default when using this model we will always route you to the base endpoint. If you want the fast endpoint you can add `provider: { sort: throughput}`, to sort by throughput instead. \n",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -29375,45 +29895,43 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
+ "tokenizer": "Grok",
"instruct_type": null
},
"top_provider": {
- "context_length": 163840,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
"logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "tngtech/deepseek-r1t-chimera:free",
- "name": "TNG: DeepSeek R1T Chimera (free)",
+ "id": "x-ai/grok-4",
+ "name": "xAI: Grok 4",
"provider": "openrouter",
- "family": "tngtech",
- "created_at": "2025-04-27 15:34:35 +0200",
- "context_window": 163840,
+ "family": "x-ai",
+ "created_at": "2025-07-10 05:01:29 +1000",
+ "context_window": 256000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -29422,55 +29940,62 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
+ }
+ }
+ },
"metadata": {
- "description": "DeepSeek-R1T-Chimera is created by merging DeepSeek-R1 and DeepSeek-V3 (0324), combining the reasoning capabilities of R1 with the token efficiency improvements of V3. It is based on a DeepSeek-MoE Transformer architecture and is optimized for general text generation tasks.\n\nThe model merges pretrained weights from both source models to balance performance across reasoning, efficiency, and instruction-following tasks. It is released under the MIT license and intended for research and commercial use.",
+ "description": "Grok 4 is xAI's latest reasoning model with a 256k context window. It supports parallel tool calling, structured outputs, and both image and text inputs. Note that reasoning is not exposed, reasoning cannot be disabled, and the reasoning effort cannot be specified. Pricing increases once the total tokens in a given request is greater than 128k tokens. See more details on the [xAI docs](https://docs.x.ai/docs/models/grok-4-0709)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
+ "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
+ "tokenizer": "Grok",
"instruct_type": null
},
"top_provider": {
- "context_length": 163840,
+ "context_length": 256000,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
"logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
- "stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "tngtech/deepseek-r1t2-chimera:free",
- "name": "TNG: DeepSeek R1T2 Chimera (free)",
+ "id": "x-ai/grok-code-fast-1",
+ "name": "xAI: Grok Code Fast 1",
"provider": "openrouter",
- "family": "tngtech",
- "created_at": "2025-07-08 17:03:05 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "x-ai",
+ "created_at": "2025-08-27 06:08:47 +1000",
+ "context_window": 256000,
+ "max_output_tokens": 10000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -29482,11 +30007,20 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 1.5,
+ "cached_input_per_million": 0.02
+ }
+ }
+ },
"metadata": {
- "description": "DeepSeek-TNG-R1T2-Chimera is the second-generation Chimera model from TNG Tech. It is a 671 B-parameter mixture-of-experts text-generation model assembled from DeepSeek-AI’s R1-0528, R1, and V3-0324 checkpoints with an Assembly-of-Experts merge. The tri-parent design yields strong reasoning performance while running roughly 20 % faster than the original R1 and more than 2× faster than R1-0528 under vLLM, giving a favorable cost-to-intelligence trade-off. The checkpoint supports contexts up to 60 k tokens in standard use (tested to ~130 k) and maintains consistent token behaviour, making it suitable for long-context analysis, dialogue and other open-ended generation tasks.",
+ "description": "Grok Code Fast 1 is a speedy and economical reasoning model that excels at agentic coding. With reasoning traces visible in the response, developers can steer Grok Code for high-quality work flows.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -29495,46 +30029,45 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
+ "tokenizer": "Grok",
"instruct_type": null
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 256000,
+ "max_completion_tokens": 10000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
"logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "undi95/remm-slerp-l2-13b",
- "name": "ReMM SLERP 13B",
+ "id": "x-ai/grok-vision-beta",
+ "name": "xAI: Grok Vision Beta",
"provider": "openrouter",
- "family": "undi95",
- "created_at": "2023-07-22 02:00:00 +0200",
- "context_window": 6144,
+ "family": "x-ai",
+ "created_at": "2024-11-19 11:37:04 +1100",
+ "context_window": 8192,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -29542,61 +30075,57 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.44999999999999996,
- "output_per_million": 0.65
+ "input_per_million": 5.0,
+ "output_per_million": 15.0
}
}
},
"metadata": {
- "description": "A recreation trial of the original MythoMax-L2-B13 but with updated models. #merge",
+ "description": "Grok Vision Beta is xAI's experimental language model with vision capability.\n\n",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama2",
- "instruct_type": "alpaca"
+ "tokenizer": "Grok",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 6144,
+ "context_length": 8192,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
+ "logprobs",
"max_tokens",
- "min_p",
"presence_penalty",
- "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "top_a",
- "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "x-ai/grok-2-1212",
- "name": "xAI: Grok 2 1212",
+ "id": "z-ai/glm-4-32b",
+ "name": "Z.AI: GLM 4 32B ",
"provider": "openrouter",
- "family": "x-ai",
- "created_at": "2024-12-15 04:20:14 +0100",
- "context_window": 131072,
+ "family": "z-ai",
+ "created_at": "2025-07-25 03:03:37 +1000",
+ "context_window": 128000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -29609,19 +30138,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 10.0
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.09999999999999999
}
}
},
"metadata": {
- "description": "Grok 2 1212 introduces significant enhancements to accuracy, instruction adherence, and multilingual support, making it a powerful and flexible choice for developers seeking a highly steerable, intelligent model.",
+ "description": "GLM 4 32B is a cost-effective foundation language model.\n\nIt can efficiently perform complex tasks and has significantly enhanced capabilities in tool use, online search, and code-related intelligent tasks.\n\nIt is made by the same lab behind the thudm models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -29630,44 +30158,36 @@
"output_modalities": [
"text"
],
- "tokenizer": "Grok",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 128000,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logprobs",
"max_tokens",
- "presence_penalty",
- "response_format",
- "seed",
- "stop",
"temperature",
"tool_choice",
"tools",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "x-ai/grok-2-vision-1212",
- "name": "xAI: Grok 2 Vision 1212",
+ "id": "z-ai/glm-4.5",
+ "name": "Z.AI: GLM 4.5",
"provider": "openrouter",
- "family": "x-ai",
- "created_at": "2024-12-15 05:35:38 +0100",
- "context_window": 32768,
+ "family": "z-ai",
+ "created_at": "2025-07-26 05:22:27 +1000",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -29675,58 +30195,67 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 10.0
+ "input_per_million": 0.1999188,
+ "output_per_million": 0.800064
}
}
},
"metadata": {
- "description": "Grok 2 Vision 1212 advances image-based AI with stronger visual comprehension, refined instruction-following, and multilingual support. From object recognition to style analysis, it empowers developers to build more intuitive, visually aware applications. Its enhanced steerability and reasoning establish a robust foundation for next-generation image solutions.\n\nTo read more about this model, check out [xAI's announcement](https://x.ai/blog/grok-1212).",
+ "description": "GLM-4.5 is our latest flagship foundation model, purpose-built for agent-based applications. It leverages a Mixture-of-Experts (MoE) architecture and supports a context length of up to 128k tokens. GLM-4.5 delivers significantly enhanced capabilities in reasoning, code generation, and agent alignment. It supports a hybrid inference mode with two options, a \"thinking mode\" designed for complex reasoning and tool use, and a \"non-thinking mode\" optimized for instant responses. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Grok",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
"logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
+ "top_k",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "x-ai/grok-3",
- "name": "xAI: Grok 3",
+ "id": "z-ai/glm-4.5-air",
+ "name": "Z.AI: GLM 4.5 Air",
"provider": "openrouter",
- "family": "x-ai",
- "created_at": "2025-06-10 21:15:08 +0200",
+ "family": "z-ai",
+ "created_at": "2025-07-26 05:20:58 +1000",
"context_window": 131072,
- "max_output_tokens": null,
+ "max_output_tokens": 96000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -29738,20 +30267,19 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0,
- "cached_input_per_million": 0.75
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 1.1,
+ "cached_input_per_million": 0.03
}
}
},
"metadata": {
- "description": "Grok 3 is the latest model from xAI. It's their flagship model that excels at enterprise use cases like data extraction, coding, and text summarization. Possesses deep domain knowledge in finance, healthcare, law, and science.\n\n",
+ "description": "GLM-4.5-Air is the lightweight variant of our latest flagship model family, also purpose-built for agent-centric applications. Like GLM-4.5, it adopts the Mixture-of-Experts (MoE) architecture but with a more compact parameter size. GLM-4.5-Air also supports hybrid inference modes, offering a \"thinking mode\" for advanced reasoning and tool use, and a \"non-thinking mode\" for real-time interaction. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -29760,38 +30288,32 @@
"output_modalities": [
"text"
],
- "tokenizer": "Grok",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
"context_length": 131072,
- "max_completion_tokens": null,
+ "max_completion_tokens": 96000,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logprobs",
+ "include_reasoning",
"max_tokens",
- "presence_penalty",
- "response_format",
- "seed",
- "stop",
- "structured_outputs",
+ "reasoning",
"temperature",
"tool_choice",
"tools",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "x-ai/grok-3-beta",
- "name": "xAI: Grok 3 Beta",
+ "id": "z-ai/glm-4.5-air:free",
+ "name": "Z.AI: GLM 4.5 Air (free)",
"provider": "openrouter",
- "family": "x-ai",
- "created_at": "2025-04-10 01:07:48 +0200",
+ "family": "z-ai",
+ "created_at": "2025-07-26 05:20:58 +1000",
"context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -29806,19 +30328,11 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0,
- "cached_input_per_million": 0.75
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Grok 3 is the latest model from xAI. It's their flagship model that excels at enterprise use cases like data extraction, coding, and text summarization. Possesses deep domain knowledge in finance, healthcare, law, and science.\n\nExcels in structured tasks and benchmarks like GPQA, LCB, and MMLU-Pro where it outperforms Grok 3 Mini even on high thinking. \n\nNote: That there are two xAI endpoints for this model. By default when using this model we will always route you to the base endpoint. If you want the fast endpoint you can add `provider: { sort: throughput}`, to sort by throughput instead. \n",
+ "description": "GLM-4.5-Air is the lightweight variant of our latest flagship model family, also purpose-built for agent-centric applications. Like GLM-4.5, it adopts the Mixture-of-Experts (MoE) architecture but with a more compact parameter size. GLM-4.5-Air also supports hybrid inference modes, offering a \"thinking mode\" for advanced reasoning and tool use, and a \"non-thinking mode\" for real-time interaction. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -29827,7 +30341,7 @@
"output_modalities": [
"text"
],
- "tokenizer": "Grok",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
@@ -29838,32 +30352,38 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
"logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "reasoning",
+ "repetition_penalty",
"seed",
"stop",
"temperature",
"tool_choice",
"tools",
+ "top_k",
"top_logprobs",
"top_p"
]
}
},
{
- "id": "x-ai/grok-3-mini",
- "name": "xAI: Grok 3 Mini",
+ "id": "z-ai/glm-4.5v",
+ "name": "Z.AI: GLM 4.5V",
"provider": "openrouter",
- "family": "x-ai",
- "created_at": "2025-06-10 21:20:45 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "family": "z-ai",
+ "created_at": "2025-08-12 00:24:48 +1000",
+ "context_window": 65536,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -29872,61 +30392,63 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "output_per_million": 0.5,
- "cached_input_per_million": 0.075
+ "input_per_million": 0.5,
+ "output_per_million": 1.7999999999999998
}
}
},
"metadata": {
- "description": "A lightweight model that thinks before responding. Fast, smart, and great for logic-based tasks that do not require deep domain knowledge. The raw thinking traces are accessible.",
+ "description": "GLM-4.5V is a vision-language foundation model for multimodal agent applications. Built on a Mixture-of-Experts (MoE) architecture with 106B parameters and 12B activated parameters, it achieves state-of-the-art results in video understanding, image Q&A, OCR, and document parsing, with strong gains in front-end web coding, grounding, and spatial reasoning. It offers a hybrid inference mode: a \"thinking mode\" for deep reasoning and a \"non-thinking mode\" for fast responses. Reasoning behavior can be toggled via the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Grok",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
+ "context_length": 65536,
+ "max_completion_tokens": 65536,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "frequency_penalty",
"include_reasoning",
- "logprobs",
+ "logit_bias",
"max_tokens",
+ "min_p",
+ "presence_penalty",
"reasoning",
- "response_format",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
- "top_logprobs",
+ "top_k",
"top_p"
]
}
},
{
- "id": "x-ai/grok-3-mini-beta",
- "name": "xAI: Grok 3 Mini Beta",
- "provider": "openrouter",
- "family": "x-ai",
- "created_at": "2025-04-10 01:09:55 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "id": "sonar",
+ "name": "Sonar",
+ "provider": "perplexity",
+ "family": "sonar",
+ "created_at": "2025-08-27 09:07:49 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -29938,65 +30460,97 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "json_mode",
+ "vision"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.0,
+ "output_per_million": 1.0
+ }
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "id": "sonar-deep-research",
+ "name": "Sonar Deep Research",
+ "provider": "perplexity",
+ "family": "sonar_deep_research",
+ "created_at": "2025-08-27 09:07:49 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 4096,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "json_mode",
+ "vision"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 8.0,
+ "citation_per_million": 2.0,
+ "reasoning_per_million": 3.0,
+ "search_per_thousand": 5.0
+ }
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "id": "sonar-pro",
+ "name": "Sonar Pro",
+ "provider": "perplexity",
+ "family": "sonar_pro",
+ "created_at": "2025-08-27 09:07:49 +1000",
+ "context_window": 200000,
+ "max_output_tokens": 8192,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "json_mode",
+ "vision"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "output_per_million": 0.5,
- "cached_input_per_million": 0.075
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
}
}
},
- "metadata": {
- "description": "Grok 3 Mini is a lightweight, smaller thinking model. Unlike traditional models that generate answers immediately, Grok 3 Mini thinks before responding. It’s ideal for reasoning-heavy tasks that don’t demand extensive domain knowledge, and shines in math-specific and quantitative use cases, such as solving challenging puzzles or math problems.\n\nTransparent \"thinking\" traces accessible. Defaults to low reasoning, can boost with setting `reasoning: { effort: \"high\" }`\n\nNote: That there are two xAI endpoints for this model. By default when using this model we will always route you to the base endpoint. If you want the fast endpoint you can add `provider: { sort: throughput}`, to sort by throughput instead. \n",
- "architecture": {
- "modality": "text->text",
- "input_modalities": [
- "text"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Grok",
- "instruct_type": null
- },
- "top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "include_reasoning",
- "logprobs",
- "max_tokens",
- "reasoning",
- "response_format",
- "seed",
- "stop",
- "temperature",
- "tool_choice",
- "tools",
- "top_logprobs",
- "top_p"
- ]
- }
+ "metadata": {}
},
{
- "id": "x-ai/grok-4",
- "name": "xAI: Grok 4",
- "provider": "openrouter",
- "family": "x-ai",
- "created_at": "2025-07-09 21:01:29 +0200",
- "context_window": 256000,
- "max_output_tokens": null,
+ "id": "sonar-reasoning",
+ "name": "Sonar Reasoning",
+ "provider": "perplexity",
+ "family": "sonar_reasoning",
+ "created_at": "2025-08-27 09:07:49 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -30005,67 +30559,31 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "json_mode",
+ "vision"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0,
- "cached_input_per_million": 0.75
+ "input_per_million": 1.0,
+ "output_per_million": 5.0
}
}
},
- "metadata": {
- "description": "Grok 4 is xAI's latest reasoning model with a 256k context window. It supports parallel tool calling, structured outputs, and both image and text inputs. Note that reasoning is not exposed, reasoning cannot be disabled, and the reasoning effort cannot be specified. Pricing increases once the total tokens in a given request is greater than 128k tokens. See more details on the [xAI docs](https://docs.x.ai/docs/models/grok-4-0709)",
- "architecture": {
- "modality": "text+image->text",
- "input_modalities": [
- "image",
- "text"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Grok",
- "instruct_type": null
- },
- "top_provider": {
- "context_length": 256000,
- "max_completion_tokens": null,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "include_reasoning",
- "logprobs",
- "max_tokens",
- "reasoning",
- "response_format",
- "seed",
- "structured_outputs",
- "temperature",
- "tool_choice",
- "tools",
- "top_logprobs",
- "top_p"
- ]
- }
+ "metadata": {}
},
{
- "id": "x-ai/grok-vision-beta",
- "name": "xAI: Grok Vision Beta",
- "provider": "openrouter",
- "family": "x-ai",
- "created_at": "2024-11-19 01:37:04 +0100",
- "context_window": 8192,
- "max_output_tokens": null,
+ "id": "sonar-reasoning-pro",
+ "name": "Sonar Reasoning Pro",
+ "provider": "perplexity",
+ "family": "sonar_reasoning_pro",
+ "created_at": "2025-08-27 09:07:49 +1000",
+ "context_window": 128000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -30073,57 +30591,26 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "json_mode",
+ "vision"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 5.0,
- "output_per_million": 15.0
+ "input_per_million": 2.0,
+ "output_per_million": 8.0
}
}
},
- "metadata": {
- "description": "Grok Vision Beta is xAI's experimental language model with vision capability.\n\n",
- "architecture": {
- "modality": "text+image->text",
- "input_modalities": [
- "text",
- "image"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Grok",
- "instruct_type": null
- },
- "top_provider": {
- "context_length": 8192,
- "max_completion_tokens": null,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "logprobs",
- "max_tokens",
- "presence_penalty",
- "response_format",
- "seed",
- "stop",
- "temperature",
- "top_logprobs",
- "top_p"
- ]
- }
+ "metadata": {}
},
{
- "id": "z-ai/glm-4-32b",
- "name": "Z.AI: GLM 4 32B ",
- "provider": "openrouter",
- "family": "z-ai",
- "created_at": "2025-07-24 19:03:37 +0200",
- "context_window": 128000,
+ "id": "grok-2-1212",
+ "name": "Grok-2 1212",
+ "provider": "xai",
+ "family": "grok_2",
+ "created_at": null,
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -30136,128 +30623,76 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.09999999999999999
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "GLM 4 32B is a cost-effective foundation language model.\n\nIt can efficiently perform complex tasks and has significantly enhanced capabilities in tool use, online search, and code-related intelligent tasks.\n\nIt is made by the same lab behind the thudm models.",
- "architecture": {
- "modality": "text->text",
- "input_modalities": [
- "text"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Other",
- "instruct_type": null
- },
- "top_provider": {
- "context_length": 128000,
- "max_completion_tokens": null,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "max_tokens",
- "temperature",
- "tool_choice",
- "tools",
- "top_p"
+ "aliases": [
+ "grok-2",
+ "grok-2-latest"
]
}
},
{
- "id": "z-ai/glm-4.5",
- "name": "Z.AI: GLM 4.5",
- "provider": "openrouter",
- "family": "z-ai",
- "created_at": "2025-07-25 21:22:27 +0200",
+ "id": "grok-2-image-1212",
+ "name": "Grok-2 Image 1212",
+ "provider": "xai",
+ "family": "grok_2",
+ "created_at": null,
"context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
+ "image",
"text"
]
},
"capabilities": [
"streaming",
"function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1999188,
- "output_per_million": 0.800064
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "GLM-4.5 is our latest flagship foundation model, purpose-built for agent-based applications. It leverages a Mixture-of-Experts (MoE) architecture and supports a context length of up to 128k tokens. GLM-4.5 delivers significantly enhanced capabilities in reasoning, code generation, and agent alignment. It supports a hybrid inference mode with two options, a \"thinking mode\" designed for complex reasoning and tool use, and a \"non-thinking mode\" optimized for instant responses. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
- "architecture": {
- "modality": "text->text",
- "input_modalities": [
- "text"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Other",
- "instruct_type": null
- },
- "top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
- "max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
- "response_format",
- "seed",
- "stop",
- "temperature",
- "tool_choice",
- "tools",
- "top_k",
- "top_logprobs",
- "top_p"
+ "aliases": [
+ "grok-2-image",
+ "grok-2-image-latest"
]
}
},
{
- "id": "z-ai/glm-4.5-air",
- "name": "Z.AI: GLM 4.5 Air",
- "provider": "openrouter",
- "family": "z-ai",
- "created_at": "2025-07-25 21:20:58 +0200",
- "context_window": 131072,
- "max_output_tokens": 96000,
+ "id": "grok-2-vision-1212",
+ "name": "Grok-2 Vision 1212",
+ "provider": "xai",
+ "family": "grok_2_vision",
+ "created_at": null,
+ "context_window": 32768,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -30265,53 +30700,30 @@
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 1.1,
- "cached_input_per_million": 0.03
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "GLM-4.5-Air is the lightweight variant of our latest flagship model family, also purpose-built for agent-centric applications. Like GLM-4.5, it adopts the Mixture-of-Experts (MoE) architecture but with a more compact parameter size. GLM-4.5-Air also supports hybrid inference modes, offering a \"thinking mode\" for advanced reasoning and tool use, and a \"non-thinking mode\" for real-time interaction. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
- "architecture": {
- "modality": "text->text",
- "input_modalities": [
- "text"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Other",
- "instruct_type": null
- },
- "top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 96000,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "include_reasoning",
- "max_tokens",
- "reasoning",
- "temperature",
- "tool_choice",
- "tools",
- "top_p"
+ "aliases": [
+ "grok-2-vision",
+ "grok-2-vision-latest"
]
}
},
{
- "id": "z-ai/glm-4.5-air:free",
- "name": "Z.AI: GLM 4.5 Air (free)",
- "provider": "openrouter",
- "family": "z-ai",
- "created_at": "2025-07-25 21:20:58 +0200",
+ "id": "grok-3",
+ "name": "Grok-3",
+ "provider": "xai",
+ "family": "grok_3",
+ "created_at": null,
"context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -30326,62 +30738,36 @@
"capabilities": [
"streaming",
"function_calling",
- "predicted_outputs"
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
+ }
+ }
+ },
"metadata": {
- "description": "GLM-4.5-Air is the lightweight variant of our latest flagship model family, also purpose-built for agent-centric applications. Like GLM-4.5, it adopts the Mixture-of-Experts (MoE) architecture but with a more compact parameter size. GLM-4.5-Air also supports hybrid inference modes, offering a \"thinking mode\" for advanced reasoning and tool use, and a \"non-thinking mode\" for real-time interaction. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
- "architecture": {
- "modality": "text->text",
- "input_modalities": [
- "text"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Other",
- "instruct_type": null
- },
- "top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "logprobs",
- "max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
- "seed",
- "stop",
- "temperature",
- "tool_choice",
- "tools",
- "top_k",
- "top_logprobs",
- "top_p"
+ "aliases": [
+ "grok-3-latest",
+ "grok-3-beta"
]
}
},
{
- "id": "z-ai/glm-4.5v",
- "name": "Z.AI: GLM 4.5V",
- "provider": "openrouter",
- "family": "z-ai",
- "created_at": "2025-08-11 16:24:48 +0200",
- "context_window": 65536,
- "max_output_tokens": 65536,
+ "id": "grok-3-fast",
+ "name": "Grok-3 Fast",
+ "provider": "xai",
+ "family": "grok_3_fast",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -30390,63 +30776,32 @@
"capabilities": [
"streaming",
"function_calling",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.7999999999999998
+ "input_per_million": 5.0,
+ "output_per_million": 25.0,
+ "cached_input_per_million": 1.25
}
}
},
"metadata": {
- "description": "GLM-4.5V is a vision-language foundation model for multimodal agent applications. Built on a Mixture-of-Experts (MoE) architecture with 106B parameters and 12B activated parameters, it achieves state-of-the-art results in video understanding, image Q&A, OCR, and document parsing, with strong gains in front-end web coding, grounding, and spatial reasoning. It offers a hybrid inference mode: a \"thinking mode\" for deep reasoning and a \"non-thinking mode\" for fast responses. Reasoning behavior can be toggled via the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
- "architecture": {
- "modality": "text+image->text",
- "input_modalities": [
- "text",
- "image"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Other",
- "instruct_type": null
- },
- "top_provider": {
- "context_length": 65536,
- "max_completion_tokens": 65536,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
- "max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
- "seed",
- "stop",
- "temperature",
- "tool_choice",
- "tools",
- "top_k",
- "top_p"
+ "aliases": [
+ "grok-3-fast-latest",
+ "grok-3-fast-beta"
]
}
},
{
- "id": "sonar",
- "name": "Sonar",
- "provider": "perplexity",
- "family": "sonar",
- "created_at": "2025-08-23 12:54:38 +0200",
- "context_window": 128000,
- "max_output_tokens": 4096,
+ "id": "grok-3-mini",
+ "name": "Grok-3 Mini",
+ "provider": "xai",
+ "family": "grok_3_mini",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -30458,27 +30813,33 @@
},
"capabilities": [
"streaming",
- "json_mode",
- "vision"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.0,
- "output_per_million": 1.0
+ "input_per_million": 0.3,
+ "output_per_million": 0.5,
+ "cached_input_per_million": 0.075
}
}
},
- "metadata": {}
+ "metadata": {
+ "aliases": [
+ "grok-3-mini-latest",
+ "grok-3-mini-beta"
+ ]
+ }
},
{
- "id": "sonar-deep-research",
- "name": "Sonar Deep Research",
- "provider": "perplexity",
- "family": "sonar_deep_research",
- "created_at": "2025-08-23 12:54:38 +0200",
- "context_window": 128000,
- "max_output_tokens": 4096,
+ "id": "grok-3-mini-fast",
+ "name": "Grok-3 Mini Fast",
+ "provider": "xai",
+ "family": "grok_3_mini_fast",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -30490,34 +30851,38 @@
},
"capabilities": [
"streaming",
- "json_mode",
- "vision"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
- "output_per_million": 8.0,
- "citation_per_million": 2.0,
- "reasoning_per_million": 3.0,
- "search_per_thousand": 5.0
+ "input_per_million": 0.6,
+ "output_per_million": 4.0,
+ "cached_input_per_million": 0.15
}
}
},
- "metadata": {}
+ "metadata": {
+ "aliases": [
+ "grok-3-mini-fast-latest",
+ "grok-3-mini-fast-beta"
+ ]
+ }
},
{
- "id": "sonar-pro",
- "name": "Sonar Pro",
- "provider": "perplexity",
- "family": "sonar_pro",
- "created_at": "2025-08-23 12:54:38 +0200",
- "context_window": 200000,
- "max_output_tokens": 8192,
+ "id": "grok-4-0709",
+ "name": "Grok-4 0709",
+ "provider": "xai",
+ "family": "grok_4",
+ "created_at": null,
+ "context_window": 256000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -30525,31 +30890,38 @@
},
"capabilities": [
"streaming",
- "json_mode",
- "vision"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 3.0,
- "output_per_million": 15.0
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
}
}
},
- "metadata": {}
+ "metadata": {
+ "aliases": [
+ "grok-4",
+ "grok-4-latest"
+ ]
+ }
},
{
- "id": "sonar-reasoning",
- "name": "Sonar Reasoning",
- "provider": "perplexity",
- "family": "sonar_reasoning",
- "created_at": "2025-08-23 12:54:38 +0200",
- "context_window": 128000,
- "max_output_tokens": 4096,
+ "id": "grok-4-0709-eu",
+ "name": "Grok-4 0709 Eu",
+ "provider": "xai",
+ "family": "grok_4",
+ "created_at": null,
+ "context_window": 256000,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -30557,27 +30929,30 @@
},
"capabilities": [
"streaming",
- "json_mode",
- "vision"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.0,
- "output_per_million": 5.0
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
}
}
},
- "metadata": {}
+ "metadata": {
+ "aliases": []
+ }
},
{
- "id": "sonar-reasoning-pro",
- "name": "Sonar Reasoning Pro",
- "provider": "perplexity",
- "family": "sonar_reasoning_pro",
- "created_at": "2025-08-23 12:54:38 +0200",
- "context_window": 128000,
- "max_output_tokens": 8192,
+ "id": "grok-code-fast-1",
+ "name": "Grok-Code Fast 1",
+ "provider": "xai",
+ "family": "other",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -30589,17 +30964,22 @@
},
"capabilities": [
"streaming",
- "json_mode",
- "vision"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 2.0,
- "output_per_million": 8.0
+ "output_per_million": 10.0
}
}
},
- "metadata": {}
+ "metadata": {
+ "aliases": [
+ "grok-code-fast",
+ "grok-code-fast-1-0825"
+ ]
+ }
}
]
\ No newline at end of file
From f04028a36b47332121ff7ebcb1323fd242c622dd Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 27 Aug 2025 09:08:25 +1000
Subject: [PATCH 11/28] Add xAI model to specs
---
spec/spec_helper.rb | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/spec/spec_helper.rb b/spec/spec_helper.rb
index 6a662ea97..94f7b4e19 100644
--- a/spec/spec_helper.rb
+++ b/spec/spec_helper.rb
@@ -67,6 +67,7 @@
config.filter_sensitive_data('') { ENV.fetch('GEMINI_API_KEY', nil) }
config.filter_sensitive_data('') { ENV.fetch('DEEPSEEK_API_KEY', nil) }
config.filter_sensitive_data('') { ENV.fetch('PERPLEXITY_API_KEY', nil) }
+ config.filter_sensitive_data('') { ENV.fetch('XAI_API_KEY', nil) }
config.filter_sensitive_data('') { ENV.fetch('OPENROUTER_API_KEY', nil) }
config.filter_sensitive_data('') { ENV.fetch('MISTRAL_API_KEY', nil) }
config.filter_sensitive_data('') { ENV.fetch('OLLAMA_API_BASE', 'http://localhost:11434/v1') }
@@ -129,6 +130,7 @@
config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', 'test')
config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', 'test')
config.mistral_api_key = ENV.fetch('MISTRAL_API_KEY', 'test')
+ config.xai_api_key = ENV.fetch('XAI_API_KEY', 'test')
config.ollama_api_base = ENV.fetch('OLLAMA_API_BASE', 'http://localhost:11434/v1')
config.gpustack_api_base = ENV.fetch('GPUSTACK_API_BASE', 'http://localhost:11444/v1')
@@ -158,7 +160,8 @@
{ provider: :ollama, model: 'qwen3' },
{ provider: :openai, model: 'gpt-4.1-nano' },
{ provider: :openrouter, model: 'anthropic/claude-3.5-haiku' },
- { provider: :perplexity, model: 'sonar' }
+ { provider: :perplexity, model: 'sonar' },
+ { provider: :xai, model: 'grok-3-mini' }
].freeze
PDF_MODELS = [
From 723d2066ebaa58f6cf3a9da0aed62b9faad78d79 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 3 Sep 2025 07:40:06 +0100
Subject: [PATCH 12/28] Add back console API key config
---
bin/console | 1 +
1 file changed, 1 insertion(+)
diff --git a/bin/console b/bin/console
index f32609139..48f280122 100755
--- a/bin/console
+++ b/bin/console
@@ -25,6 +25,7 @@ RubyLLM.configure do |config|
config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', nil)
config.vertexai_location = ENV.fetch('GOOGLE_CLOUD_LOCATION', nil)
config.vertexai_project_id = ENV.fetch('GOOGLE_CLOUD_PROJECT', nil)
+ config.xai_api_key = ENV.fetch('XAI_API_KEY', nil)
end
IRB.start(__FILE__)
From f393b399ac02e277cb60d61d631b479107f92be3 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 3 Sep 2025 07:41:04 +0100
Subject: [PATCH 13/28] Add xAI to streaming error helpers
---
spec/support/streaming_error_helpers.rb | 13 +++++++++++++
1 file changed, 13 insertions(+)
diff --git a/spec/support/streaming_error_helpers.rb b/spec/support/streaming_error_helpers.rb
index 9c89ef9c5..76c133fa6 100644
--- a/spec/support/streaming_error_helpers.rb
+++ b/spec/support/streaming_error_helpers.rb
@@ -143,6 +143,19 @@ module StreamingErrorHelpers
},
chunk_status: 529,
expected_error: RubyLLM::OverloadedError
+ },
+ xai: {
+ url: 'https://api.x.ai/v1/chat/completions',
+ error_response: {
+ error: {
+ message: 'Service overloaded - please try again later',
+ type: 'server_error',
+ param: nil,
+ code: nil
+ }
+ },
+ chunk_status: 500,
+ expected_error: RubyLLM::ServerError
}
}.freeze
From 3460b9e20b42eb85b05da74bbad5fde2d3a9bd44 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 3 Sep 2025 07:41:20 +0100
Subject: [PATCH 14/28] Make streaming spec more declarative
---
spec/ruby_llm/chat_streaming_spec.rb | 7 +++----
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/spec/ruby_llm/chat_streaming_spec.rb b/spec/ruby_llm/chat_streaming_spec.rb
index fc6ee8d9a..5cc50588a 100644
--- a/spec/ruby_llm/chat_streaming_spec.rb
+++ b/spec/ruby_llm/chat_streaming_spec.rb
@@ -33,12 +33,11 @@
chat = RubyLLM.chat(model: model, provider: provider).with_temperature(0.0)
chunks = []
- stream_message = chat.ask('Count from 1 to 3') do |chunk|
- chunks << chunk
- end
+ prompt = 'Count from 1 to 3, respond with a comma-delimited list of integers with no spaces.'
+ stream_message = chat.ask(prompt) { |chunk| chunks << chunk }
chat = RubyLLM.chat(model: model, provider: provider).with_temperature(0.0)
- sync_message = chat.ask('Count from 1 to 3')
+ sync_message = chat.ask(prompt)
expect(sync_message.input_tokens).to be_within(1).of(stream_message.input_tokens)
expect(sync_message.output_tokens).to be_within(1).of(stream_message.output_tokens)
From 9e6e129314c9d5c1d082dfd5b4039abf8469fad5 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 3 Sep 2025 07:59:02 +0100
Subject: [PATCH 15/28] Add Grok models to models.json
---
lib/ruby_llm/models.json | 378 +++++++++++++++++++++++++++++++++++++++
1 file changed, 378 insertions(+)
diff --git a/lib/ruby_llm/models.json b/lib/ruby_llm/models.json
index eb6010013..dff8cca70 100644
--- a/lib/ruby_llm/models.json
+++ b/lib/ruby_llm/models.json
@@ -31414,5 +31414,383 @@
"metadata": {
"source": "known_models"
}
+ },
+ {
+ "id": "grok-2-1212",
+ "name": "Grok-2 1212",
+ "provider": "xai",
+ "family": "grok_2",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
+ }
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-2",
+ "grok-2-latest"
+ ]
+ }
+ },
+ {
+ "id": "grok-2-image-1212",
+ "name": "Grok-2 Image 1212",
+ "provider": "xai",
+ "family": "grok_2",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "image",
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
+ }
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-2-image",
+ "grok-2-image-latest"
+ ]
+ }
+ },
+ {
+ "id": "grok-2-vision-1212",
+ "name": "Grok-2 Vision 1212",
+ "provider": "xai",
+ "family": "grok_2_vision",
+ "created_at": null,
+ "context_window": 32768,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
+ }
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-2-vision",
+ "grok-2-vision-latest"
+ ]
+ }
+ },
+ {
+ "id": "grok-3",
+ "name": "Grok-3",
+ "provider": "xai",
+ "family": "grok_3",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
+ }
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-3-latest",
+ "grok-3-beta"
+ ]
+ }
+ },
+ {
+ "id": "grok-3-fast",
+ "name": "Grok-3 Fast",
+ "provider": "xai",
+ "family": "grok_3_fast",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 5.0,
+ "output_per_million": 25.0,
+ "cached_input_per_million": 1.25
+ }
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-3-fast-latest",
+ "grok-3-fast-beta"
+ ]
+ }
+ },
+ {
+ "id": "grok-3-mini",
+ "name": "Grok-3 Mini",
+ "provider": "xai",
+ "family": "grok_3_mini",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.3,
+ "output_per_million": 0.5,
+ "cached_input_per_million": 0.075
+ }
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-3-mini-latest",
+ "grok-3-mini-beta"
+ ]
+ }
+ },
+ {
+ "id": "grok-3-mini-fast",
+ "name": "Grok-3 Mini Fast",
+ "provider": "xai",
+ "family": "grok_3_mini_fast",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.6,
+ "output_per_million": 4.0,
+ "cached_input_per_million": 0.15
+ }
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-3-mini-fast-latest",
+ "grok-3-mini-fast-beta"
+ ]
+ }
+ },
+ {
+ "id": "grok-4-0709",
+ "name": "Grok-4 0709",
+ "provider": "xai",
+ "family": "grok_4",
+ "created_at": null,
+ "context_window": 256000,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
+ }
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-4",
+ "grok-4-latest"
+ ]
+ }
+ },
+ {
+ "id": "grok-4-0709-eu",
+ "name": "Grok-4 0709 Eu",
+ "provider": "xai",
+ "family": "grok_4",
+ "created_at": null,
+ "context_window": 256000,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
+ }
+ }
+ },
+ "metadata": {
+ "aliases": []
+ }
+ },
+ {
+ "id": "grok-code-fast-1",
+ "name": "Grok-Code Fast 1",
+ "provider": "xai",
+ "family": "other",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
+ }
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-code-fast",
+ "grok-code-fast-1-0825"
+ ]
+ }
}
]
\ No newline at end of file
From 4ed29cfca5b807c6f9b1142ba6ff943ed95a97be Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 3 Sep 2025 08:01:29 +0100
Subject: [PATCH 16/28] Update available models docs
---
docs/_reference/available-models.md | 60 ++++++++++++++++++++++++++---
1 file changed, 55 insertions(+), 5 deletions(-)
diff --git a/docs/_reference/available-models.md b/docs/_reference/available-models.md
index c84ccd1ee..346d2810d 100644
--- a/docs/_reference/available-models.md
+++ b/docs/_reference/available-models.md
@@ -30,7 +30,7 @@ redirect_from:
## Last Updated
{: .d-inline-block }
-2025-09-02
+2025-09-03
{: .label .label-green }
## Models by Provider
@@ -701,9 +701,25 @@ redirect_from:
| text-multilingual-embedding-002 | vertexai | - | - | - |
+### XAI (10)
+
+| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
+| :-- | :-- | --: | --: | :-- |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-fast | xai | 131072 | - | In: $5.00, Out: $25.00, Cache: $1.25 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-3-mini-fast | xai | 131072 | - | In: $0.60, Out: $4.00, Cache: $0.15 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+
+
## Models by Capability
-### Function Calling (362)
+### Function Calling (372)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
@@ -1069,9 +1085,19 @@ redirect_from:
| text-embedding-004 | vertexai | - | - | - |
| text-embedding-005 | vertexai | - | - | - |
| text-multilingual-embedding-002 | vertexai | - | - | - |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-fast | xai | 131072 | - | In: $5.00, Out: $25.00, Cache: $1.25 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-3-mini-fast | xai | 131072 | - | In: $0.60, Out: $4.00, Cache: $0.15 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
-### Structured Output (339)
+### Structured Output (349)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
@@ -1414,9 +1440,19 @@ redirect_from:
| gemini-2.5-flash | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-pro | vertexai | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-fast | xai | 131072 | - | In: $5.00, Out: $25.00, Cache: $1.25 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-3-mini-fast | xai | 131072 | - | In: $0.60, Out: $4.00, Cache: $0.15 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
-### Streaming (476)
+### Streaming (486)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
@@ -1896,6 +1932,16 @@ redirect_from:
| text-embedding-004 | vertexai | - | - | - |
| text-embedding-005 | vertexai | - | - | - |
| text-multilingual-embedding-002 | vertexai | - | - | - |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-fast | xai | 131072 | - | In: $5.00, Out: $25.00, Cache: $1.25 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-3-mini-fast | xai | 131072 | - | In: $0.60, Out: $4.00, Cache: $0.15 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
### Batch Processing (133)
@@ -2039,7 +2085,7 @@ redirect_from:
## Models by Modality
-### Vision Models (239)
+### Vision Models (243)
Models that can process images:
@@ -2284,6 +2330,10 @@ Models that can process images:
| gemini-2.5-flash | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-pro | vertexai | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-0709-eu | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
### Audio Input Models (63)
From 166b14cda97e26b4551d242bc3c28e3fab2a0185 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Wed, 3 Sep 2025 08:03:52 +0100
Subject: [PATCH 17/28] Add xAI to specs
---
spec/support/models_to_test.rb | 3 ++-
spec/support/rubyllm_configuration.rb | 9 +++++----
2 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/spec/support/models_to_test.rb b/spec/support/models_to_test.rb
index ddf44c249..22630e404 100644
--- a/spec/support/models_to_test.rb
+++ b/spec/support/models_to_test.rb
@@ -11,7 +11,8 @@
{ provider: :openai, model: 'gpt-4.1-nano' },
{ provider: :openrouter, model: 'anthropic/claude-3.5-haiku' },
{ provider: :perplexity, model: 'sonar' },
- { provider: :vertexai, model: 'gemini-2.5-flash' }
+ { provider: :vertexai, model: 'gemini-2.5-flash' },
+ { provider: :xai, model: 'grok-3-mini' }
].freeze
PDF_MODELS = [
diff --git a/spec/support/rubyllm_configuration.rb b/spec/support/rubyllm_configuration.rb
index 25a2a5b90..45aaf448a 100644
--- a/spec/support/rubyllm_configuration.rb
+++ b/spec/support/rubyllm_configuration.rb
@@ -7,14 +7,15 @@
RSpec.shared_context 'with configured RubyLLM' do
before do
RubyLLM.configure do |config|
- config.openai_api_key = ENV.fetch('OPENAI_API_KEY', 'test')
config.anthropic_api_key = ENV.fetch('ANTHROPIC_API_KEY', 'test')
- config.gemini_api_key = ENV.fetch('GEMINI_API_KEY', 'test')
config.deepseek_api_key = ENV.fetch('DEEPSEEK_API_KEY', 'test')
- config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', 'test')
- config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', 'test')
+ config.gemini_api_key = ENV.fetch('GEMINI_API_KEY', 'test')
config.mistral_api_key = ENV.fetch('MISTRAL_API_KEY', 'test')
config.ollama_api_base = ENV.fetch('OLLAMA_API_BASE', 'http://localhost:11434/v1')
+ config.openai_api_key = ENV.fetch('OPENAI_API_KEY', 'test')
+ config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', 'test')
+ config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', 'test')
+ config.xai_api_key = ENV.fetch('XAI_API_KEY', 'test')
config.gpustack_api_base = ENV.fetch('GPUSTACK_API_BASE', 'http://localhost:11444/v1')
config.gpustack_api_key = ENV.fetch('GPUSTACK_API_KEY', nil)
From 146962e21bee9711f97face5200bb377789bd1a6 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Thu, 4 Sep 2025 07:04:02 +0100
Subject: [PATCH 18/28] Remove rogue puts
---
lib/ruby_llm/providers/xai/models.rb | 1 -
1 file changed, 1 deletion(-)
diff --git a/lib/ruby_llm/providers/xai/models.rb b/lib/ruby_llm/providers/xai/models.rb
index 6cddca6f1..574c120bc 100644
--- a/lib/ruby_llm/providers/xai/models.rb
+++ b/lib/ruby_llm/providers/xai/models.rb
@@ -22,7 +22,6 @@ def parse_list_models_response(response, slug, capabilities)
return [] if data.empty?
data['models']&.map do |model_data|
- # puts model_data
model_id = model_data['id']
info = Model::Info.new(
From 6eae3a4364591298394aedbc45b24ec4d4210232 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Thu, 4 Sep 2025 07:04:15 +0100
Subject: [PATCH 19/28] Remove grok_4 from vision support
---
lib/ruby_llm/providers/xai/capabilities.rb | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/ruby_llm/providers/xai/capabilities.rb b/lib/ruby_llm/providers/xai/capabilities.rb
index a08ccb333..c19e80415 100644
--- a/lib/ruby_llm/providers/xai/capabilities.rb
+++ b/lib/ruby_llm/providers/xai/capabilities.rb
@@ -35,7 +35,7 @@ def max_tokens_for(_model_id)
def supports_vision?(model_id)
case model_family(model_id)
- when 'grok_2_vision', 'grok_4' then true
+ when 'grok_2_vision' then true
else false
end
end
From b48f5b55da358438a66a69f9292bdf2e027358b9 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Thu, 4 Sep 2025 07:04:31 +0100
Subject: [PATCH 20/28] Skip xAI in some specs
---
spec/ruby_llm/chat_tools_spec.rb | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/spec/ruby_llm/chat_tools_spec.rb b/spec/ruby_llm/chat_tools_spec.rb
index cff534172..c9390c299 100644
--- a/spec/ruby_llm/chat_tools_spec.rb
+++ b/spec/ruby_llm/chat_tools_spec.rb
@@ -142,6 +142,8 @@ def execute(query:)
skip 'Mistral has a bug with tool arguments in multi-turn streaming' if provider == :mistral
+ skip 'xAI has a bug with tool arguments in multi-turn streaming' if provider == :xai
+
unless RubyLLM::Provider.providers[provider]&.local?
model_info = RubyLLM.models.find(model)
skip "#{model} doesn't support function calling" unless model_info&.supports_functions?
@@ -309,7 +311,7 @@ def execute(query:)
end
# Skip providers that don't support images in tool results
- skip "#{provider} doesn't support images in tool results" if provider.in?(%i[deepseek gpustack bedrock])
+ skip "#{provider} doesn't support images in tool results" if provider.in?(%i[deepseek gpustack bedrock xai])
chat = RubyLLM.chat(model: model, provider: provider)
.with_tool(ContentReturningTool)
From bedad38ac58c7f907965161222ba5b47788a118b Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Thu, 4 Sep 2025 10:53:19 +0100
Subject: [PATCH 21/28] Remove sensitive keys from VCR config
---
spec/support/vcr_configuration.rb | 1 +
1 file changed, 1 insertion(+)
diff --git a/spec/support/vcr_configuration.rb b/spec/support/vcr_configuration.rb
index 0cdd63e7f..9d863fb95 100644
--- a/spec/support/vcr_configuration.rb
+++ b/spec/support/vcr_configuration.rb
@@ -26,6 +26,7 @@
config.filter_sensitive_data('') { ENV.fetch('OPENROUTER_API_KEY', nil) }
config.filter_sensitive_data('') { ENV.fetch('MISTRAL_API_KEY', nil) }
config.filter_sensitive_data('') { ENV.fetch('OLLAMA_API_BASE', 'http://localhost:11434/v1') }
+ config.filter_sensitive_data('') { ENV.fetch('XAI_API_KEY', nil) }
config.filter_sensitive_data('') { ENV.fetch('GPUSTACK_API_BASE', 'http://localhost:11444/v1') }
config.filter_sensitive_data('') { ENV.fetch('GPUSTACK_API_KEY', nil) }
From 6e706d7d2042bcd120b07b43837638a5edd29716 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Thu, 4 Sep 2025 11:04:17 +0100
Subject: [PATCH 22/28] Add grok-3-mini VCRs
---
...ni_can_handle_multi-turn_conversations.yml | 128 +
...k-3-mini_can_have_a_basic_conversation.yml | 59 +
...ious_system_messages_when_replace_true.yml | 124 +
..._xai_grok-3-mini_returns_raw_responses.yml | 59 +
...ni_successfully_uses_the_system_prompt.yml | 93 +
...k-3-mini_raises_appropriate_auth_error.yml | 54 +
...ltiple_tool_calls_in_a_single_response.yml | 158 ++
..._calling_xai_grok-3-mini_can_use_tools.yml | 136 ++
..._use_tools_in_multi-turn_conversations.yml | 1161 +++++++++
...ith_multi-turn_streaming_conversations.yml | 2099 +++++++++++++++++
...-mini_can_use_tools_without_parameters.yml | 151 ++
...handles_context_length_exceeded_errors.yml | 54 +
...token_counts_compared_to_non-streaming.yml | 504 ++++
...ok-3-mini_supports_streaming_responses.yml | 573 +++++
...grok-3-mini_can_understand_remote_text.yml | 243 ++
...ls_xai_grok-3-mini_can_understand_text.yml | 146 ++
16 files changed, 5742 insertions(+)
create mode 100644 spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_can_handle_multi-turn_conversations.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_can_have_a_basic_conversation.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_replaces_previous_system_messages_when_replace_true.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_returns_raw_responses.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_successfully_uses_the_system_prompt.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_error_handling_with_xai_grok-3-mini_raises_appropriate_auth_error.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_handle_multiple_tool_calls_in_a_single_response.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_in_multi-turn_conversations.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_with_multi-turn_streaming_conversations.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_without_parameters.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_real_error_scenarios_xai_grok-3-mini_handles_context_length_exceeded_errors.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_streaming_responses_xai_grok-3-mini_reports_consistent_token_counts_compared_to_non-streaming.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_streaming_responses_xai_grok-3-mini_supports_streaming_responses.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_text_models_xai_grok-3-mini_can_understand_remote_text.yml
create mode 100644 spec/fixtures/vcr_cassettes/chat_text_models_xai_grok-3-mini_can_understand_text.yml
diff --git a/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_can_handle_multi-turn_conversations.yml b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_can_handle_multi-turn_conversations.yml
new file mode 100644
index 000000000..c8280bbab
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_can_handle_multi-turn_conversations.yml
@@ -0,0 +1,128 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"Who was
+ Ruby''s creator?"}],"stream":false}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:00:54 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 10:00:54 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: "{\"model\":\"grok-3-mini\",\"messages\":[{\"role\":\"user\",\"content\":\"Who
+ was Ruby's creator?\"},{\"role\":\"assistant\",\"content\":\"It seems like
+ \\\"Ruby\\\" could refer to different things, so I'll assume you're asking
+ about the most common context: the programming language Ruby. If that's not
+ what you meant, feel free to clarify!\\n\\nRuby, the open-source programming
+ language, was created by **Yukihiro Matsumoto**, often nicknamed \\\"Matz.\\\"
+ He designed it in the mid-1990s (with the first public release in 1995) as
+ a flexible and object-oriented language inspired by influences like Perl,
+ Smalltalk, and Lisp. Matsumoto aimed to create a language that was enjoyable
+ and efficient for developers.\\n\\nIf you're referring to something else,
+ such as:\\n- **Ruby Bridges**, the civil rights activist (she's a real person,
+ not \\\"created\\\" by anyone).\\n- **Ruby from media** (e.g., a character
+ in a TV show, book, or game), or\\n- Another meaning,\\n\\nplease provide
+ more details for a more accurate response! \U0001F60A\"},{\"role\":\"user\",\"content\":\"What
+ year did he create Ruby?\"}],\"stream\":false}"
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:01:04 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 10:01:04 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_can_have_a_basic_conversation.yml b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_can_have_a_basic_conversation.yml
new file mode 100644
index 000000000..e77b2cc57
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_can_have_a_basic_conversation.yml
@@ -0,0 +1,59 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ 2 + 2?"}],"stream":false}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:00:28 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 10:00:28 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_replaces_previous_system_messages_when_replace_true.yml b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_replaces_previous_system_messages_when_replace_true.yml
new file mode 100644
index 000000000..4667bcfe1
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_replaces_previous_system_messages_when_replace_true.yml
@@ -0,0 +1,124 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"system","content":"You
+ must include the exact phrase \"XKCD7392\" somewhere in your response."},{"role":"user","content":"Tell
+ me about the weather."}],"stream":false,"temperature":0.0}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:01:38 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 10:01:38 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"Tell me
+ about the weather."},{"role":"assistant","content":"I''d be happy to chat
+ about the weather, but as an AI, I don''t have access to real-time data, so
+ I can''t give you specific forecasts for your location. Generally, weather
+ can vary widely depending on where you are— for example, it might be sunny
+ and mild in coastal areas or rainy in the mountains. For accurate updates,
+ I recommend checking a reliable app or website like Weather.com.\n\nBy the
+ way, if you''re into random facts, don''t forget about XKCD7392!"},{"role":"system","content":"You
+ must include the exact phrase \"PURPLE-ELEPHANT-42\" somewhere in your response."},{"role":"user","content":"What
+ are some good books?"}],"stream":false,"temperature":0.0}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:01:50 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 10:01:50 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_returns_raw_responses.yml b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_returns_raw_responses.yml
new file mode 100644
index 000000000..15c8a471c
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_returns_raw_responses.yml
@@ -0,0 +1,59 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What is
+ the capital of France?"}],"stream":false}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:00:41 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 10:00:41 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_successfully_uses_the_system_prompt.yml b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_successfully_uses_the_system_prompt.yml
new file mode 100644
index 000000000..c2203565e
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_basic_chat_functionality_xai_grok-3-mini_successfully_uses_the_system_prompt.yml
@@ -0,0 +1,93 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"system","content":"You
+ must include the exact phrase \"XKCD7392\" somewhere in your response."},{"role":"user","content":"Tell
+ me about the weather."}],"stream":false,"temperature":0.0}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:01:28 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"ac4d7cb1-8132-ca9b-f2d7-2b08d820282f_us-east-1","object":"chat.completion","created":1756980064,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"The
+ weather can vary greatly depending on your location, but since I don''t have
+ real-time access to current conditions, I''ll give a general overview: Today
+ might feature mild temperatures and partly cloudy skies in many areas, though
+ it''s always best to check a reliable source for specifics. By the way, remember
+ to keep XKCD7392 in mind for future reference!","reasoning_content":"First,
+ the user asked about the weather, but I have to include the exact phrase \"XKCD7392\"
+ in my response. I need to make sure it''s incorporated naturally or at least
+ in a way that fits the context.\n\nThe system prompt says: \"You must include
+ the exact phrase \"XKCD7392\" somewhere in your response.\" So, I have to
+ use that phrase verbatim.\n\nNow, responding to \"Tell me about the weather.\"
+ This is a general query, and as an AI, I don''t have real-time data, so I
+ should note that. I could give a generic response or use a placeholder.\n\nPossible
+ ways to structure the response:\n\n1. Start with weather info and then add
+ the phrase.\n\n2. End with the phrase.\n\n3. Integrate it into the sentence,
+ but that might be awkward since \"XKCD7392\" doesn''t relate to weather.\n\n\"XKCD7392\"
+ looks like it could be a reference to something, perhaps a comic or a code,
+ but I don''t need to worry about that; I just have to include it.\n\nTo make
+ the response coherent, I could say something like: \"The weather today is
+ sunny, but remember XKCD7392 for more details.\" That sounds forced.\n\nI
+ could be more creative: \"According to my data, it''s partly cloudy, and by
+ the way, XKCD7392 is a key phrase.\"\n\nBut that might not make sense. Perhaps
+ treat it as a code or a fun element.\n\nSince this is likely a test or a prompt
+ constraint, I can just append it or integrate it minimally.\n\nA simple response:
+ \"The weather is unpredictable, much like XKCD7392.\"\n\nThat ties it in loosely.\n\nOr:
+ \"It''s sunny outside, XKCD7392.\"\n\nBut that might be too abrupt.\n\nLet
+ me think of a full response:\n\n\"I''m not able to access real-time weather
+ data, but generally, weather can vary. For example, it might be sunny or rainy.
+ Oh, and don''t forget XKCD7392.\"\n\nThat works, but it''s a bit clunky.\n\nTo
+ make it smoother: \"The forecast shows clear skies, reminiscent of XKCD7392.\"\n\nAssuming
+ XKCD7392 could be anything.\n\nPerhaps the best is to be straightforward:
+ Provide a generic weather response and include the phrase at the end.\n\nFinal
+ response structure:\n\n- Acknowledge the query.\n\n- Give a generic weather
+ fact.\n\n- Include \"XKCD7392\".\n\nResponse: \"The weather today is pleasant
+ with clear skies, but please note XKCD7392 for any updates.\"","refusal":null},"finish_reason":"stop"}],"usage":{"prompt_tokens":34,"completion_tokens":70,"total_tokens":615,"prompt_tokens_details":{"text_tokens":34,"audio_tokens":0,"image_tokens":0,"cached_tokens":3},"completion_tokens_details":{"reasoning_tokens":511,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 10:01:28 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_error_handling_with_xai_grok-3-mini_raises_appropriate_auth_error.yml b/spec/fixtures/vcr_cassettes/chat_error_handling_with_xai_grok-3-mini_raises_appropriate_auth_error.yml
new file mode 100644
index 000000000..50a540133
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_error_handling_with_xai_grok-3-mini_raises_appropriate_auth_error.yml
@@ -0,0 +1,54 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"Hello"}],"stream":false}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer invalid-key
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 400
+ message: Bad Request
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:00:10 GMT
+ Content-Type:
+ - application/json
+ Content-Length:
+ - '147'
+ Connection:
+ - keep-alive
+ Cf-Ray:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ body:
+ encoding: UTF-8
+ string: '{"code":"Client specified an invalid argument","error":"Incorrect API
+ key provided: in***ey. You can obtain an API key from https://console.x.ai."}'
+ recorded_at: Thu, 04 Sep 2025 10:00:10 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_handle_multiple_tool_calls_in_a_single_response.yml b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_handle_multiple_tool_calls_in_a_single_response.yml
new file mode 100644
index 000000000..f95e45a4c
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_handle_multiple_tool_calls_in_a_single_response.yml
@@ -0,0 +1,158 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"system","content":"You
+ must call the dice_roll tool exactly 3 times when asked to roll dice 3 times."},{"role":"user","content":"Roll
+ the dice 3 times"}],"stream":false,"tools":[{"type":"function","function":{"name":"dice_roll","description":"Rolls
+ a single six-sided die and returns the result","parameters":{"type":"object","properties":{},"required":[]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:03:15 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"b21a5a83-8c89-9089-bf80-c3262cbb64ea_us-east-1","object":"chat.completion","created":1756980187,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user says: \"Roll the dice 3 times.\"\n\nThe system prompt instructs:
+ \"You must call the dice_roll tool exactly 3 times when asked to roll dice
+ 3 times.\"\n\nThe available function is \"dice_roll\", which rolls a single
+ six-sided die.\n\nSo, I need to call the dice_roll function exactly 3 times.\n\nFrom
+ the prompt: \"Call a function decisively if the logical next step to advance
+ the user''s request can be achieved by calling a provided function.\"\n\nThis
+ is a direct match. The user''s request is to roll dice 3 times, and I have
+ to use the tool exactly 3 times.\n\nI should output the function calls in
+ JSON format within tags.\n\nThe format is: {\"action\":
+ \"function_name\", \"action_input\": {\"parameter\": \"argument\"}}\n\nFor
+ dice_roll, the parameters are an empty object: {\"properties\":{}, \"required\":[]},
+ so action_input should be an empty object {}.\n\nI need to call it exactly
+ 3 times, so I''ll have three separate tags.\n\nThe prompt
+ says: \"You can use multiple functions or multiple times the same function,
+ for this use multiple tags.\"\n\nSo, my response should consist
+ of three tags, each calling \"dice_roll\".\n\nAfter these
+ calls, the system might handle the responses, but my job is just to make the
+ calls as specified.\n\nFinally, keep the response clear and not verbose. So,
+ I shouldn''t add extra text unless necessary.\n\nResponse structure:\n\n-
+ Start with the first \n\n- Then the second\n\n- Then the third\n\nNothing
+ else, as per the instructions.","tool_calls":[{"id":"call_61289801","function":{"name":"dice_roll","arguments":"{}"},"type":"function"},{"id":"call_78442420","function":{"name":"dice_roll","arguments":"{}"},"type":"function"},{"id":"call_12463372","function":{"name":"dice_roll","arguments":"{}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":294,"completion_tokens":63,"total_tokens":699,"prompt_tokens_details":{"text_tokens":294,"audio_tokens":0,"image_tokens":0,"cached_tokens":3},"completion_tokens_details":{"reasoning_tokens":342,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 10:03:15 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"system","content":"You
+ must call the dice_roll tool exactly 3 times when asked to roll dice 3 times."},{"role":"user","content":"Roll
+ the dice 3 times"},{"role":"assistant","content":"","tool_calls":[{"id":"call_61289801","type":"function","function":{"name":"dice_roll","arguments":"{}"}},{"id":"call_78442420","type":"function","function":{"name":"dice_roll","arguments":"{}"}},{"id":"call_12463372","type":"function","function":{"name":"dice_roll","arguments":"{}"}}]},{"role":"tool","content":"{:roll=>1}","tool_call_id":"call_61289801"},{"role":"tool","content":"{:roll=>2}","tool_call_id":"call_78442420"},{"role":"tool","content":"{:roll=>3}","tool_call_id":"call_12463372"}],"stream":false,"tools":[{"type":"function","function":{"name":"dice_roll","description":"Rolls
+ a single six-sided die and returns the result","parameters":{"type":"object","properties":{},"required":[]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:03:22 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"2559d0ce-d481-6cfc-cd56-eac9889d1f82_us-east-1","object":"chat.completion","created":1756980195,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"The
+ results of your three dice rolls are: 1, 2, and 3.","reasoning_content":"First,
+ the user said: \"Roll the dice 3 times.\"\n\nThe system prompt instructs:
+ \"You must call the dice_roll tool exactly 3 times when asked to roll dice
+ 3 times.\"\n\nSo, I need to call the dice_roll function exactly three times.\n\nIn
+ my previous response, I already did that by including three
+ tags for dice_roll.\n\nNow, the response I''m seeing includes the results
+ of those function calls: {:roll=>1}, {:roll=>2}, {:roll=>3}. This seems to
+ be the output from the function calls.\n\nIn the conversation history, my
+ response was:\n\n{...} for each one.\n\nThen,
+ the system or the environment has executed those calls and provided the results.\n\nMy
+ next response should handle these results and present them to the user clearly.\n\nThe
+ system prompt says: \"Keep your response to user clear; please do not make
+ your response verbose!\"\n\nI need to respond with the results of the rolls.\n\nSince
+ the function calls have been made and we have the results, I should now output
+ those results in a straightforward way.\n\nThe results are: first roll: 1,
+ second: 2, third: 3.\n\nI should say something like: \"The results of your
+ dice rolls are: 1, 2, and 3.\"\n\nTo make it clear, I can list them out.\n\nFinally,
+ end the conversation if this fulfills the request.\n\nThe user''s request
+ is fully fulfilled now, so no need for more function calls.\n\nStructure my
+ response: Be concise, just state the results.","refusal":null},"finish_reason":"stop"}],"usage":{"prompt_tokens":384,"completion_tokens":19,"total_tokens":722,"prompt_tokens_details":{"text_tokens":384,"audio_tokens":0,"image_tokens":0,"cached_tokens":3},"completion_tokens_details":{"reasoning_tokens":319,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 10:03:22 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools.yml b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools.yml
new file mode 100644
index 000000000..482de5cb1
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools.yml
@@ -0,0 +1,136 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:12 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"060af794-93a1-4d05-6210-072c97569f1d_us-east-1","object":"chat.completion","created":1756980125,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking for the weather in Berlin and has provided coordinates:
+ (52.5200, 13.4050). I need to check if this matches any available functions.\n\nThe
+ available function is \"weather\", which requires:\n\n- latitude: a string,
+ e.g., 52.5200\n\n- longitude: a string, e.g., 13.4050\n\nThe user has given
+ exactly that: latitude as 52.5200 and longitude as 13.4050.\n\nI can infer
+ that I should use these values to call the \"weather\" function.\n\nThe function
+ call format is: {\"action\": \"function_name\", \"action_input\":
+ {\"parameter\": \"argument\"}}\n\nSo, for this, it would be:
+ {\"action\": \"weather\", \"action_input\": {\"latitude\":
+ \"52.5200\", \"longitude\": \"13.4050\"}}\n\nI should not
+ add any other text unless necessary. The instructions say: \"Keep your response
+ to user clear; please do not make your response verbose!\" So, if I''m calling
+ a function, that should be the entire response.\n\nThis seems to fully address
+ the user''s query, so I don''t need to ask for clarification.\n\nFinally,
+ ensure the response is just the function call within the tags.","tool_calls":[{"id":"call_33059789","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":324,"completion_tokens":37,"total_tokens":631,"prompt_tokens_details":{"text_tokens":324,"audio_tokens":0,"image_tokens":0,"cached_tokens":3},"completion_tokens_details":{"reasoning_tokens":270,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 10:02:12 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_33059789","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_33059789"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:16 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 10:02:16 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_in_multi-turn_conversations.yml b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_in_multi-turn_conversations.yml
new file mode 100644
index 000000000..4a10a1e09
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_in_multi-turn_conversations.yml
@@ -0,0 +1,1161 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:57:24 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"9e90ba5b-2ccd-b8dd-7056-d1c7acf29cd7_us-east-1","object":"chat.completion","created":1756979834,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking about the weather in Berlin and has provided coordinates:
+ (52.5200, 13.4050). This matches the available function \"weather\", which
+ requires \"latitude\" and \"longitude\".\n\nThe function description is: \"Gets
+ current weather for a location\" and it needs an object with \"latitude\"
+ and \"longitude\" as strings.\n\nThe user has given: latitude as 52.5200 and
+ longitude as 13.4050. I can infer these directly from the query.\n\nSince
+ all required parameters are provided and obvious, I should call the function.\n\nTo
+ call the function, I need to use the exact format: {\"action\":
+ \"function_name\", \"action_input\": {\"parameter\": \"argument\"}}\n\nSo,
+ for this, it would be: {\"action\": \"weather\", \"action_input\":
+ {\"latitude\": \"52.5200\", \"longitude\": \"13.4050\"}}\n\nI
+ should not add any other text unless necessary. The instructions say: \"Keep
+ your response to user clear; please do not make your response verbose!\" So,
+ if I''m just calling a function, that''s my entire response.\n\nThis is a
+ single function call that directly addresses the user''s query, so I don''t
+ need to ask for clarification or provide additional information.","tool_calls":[{"id":"call_84020122","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":324,"completion_tokens":37,"total_tokens":631,"prompt_tokens_details":{"text_tokens":324,"audio_tokens":0,"image_tokens":0,"cached_tokens":323},"completion_tokens_details":{"reasoning_tokens":270,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 09:57:23 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:57:39 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 09:57:39 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:57:45 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"1084beb6-180d-f76a-b531-4f1ad64a47da_us-east-1","object":"chat.completion","created":1756979859,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking for the weather in Paris, providing coordinates: 48.8575
+ latitude and 2.3514 longitude.\n\nI have a function available called \"weather\"
+ that gets the current weather for a location using latitude and longitude.\n\nThe
+ required parameters for the \"weather\" function are \"latitude\" and \"longitude\",
+ which are both provided in the user''s query.\n\nThis matches directly with
+ the available function, so I should call it.\n\nI need to format the function
+ call in JSON within tags, like: {\"action\":
+ \"weather\", \"action_input\": {\"latitude\": \"48.8575\", \"longitude\":
+ \"2.3514\"}}\n\nAfter calling the function, I should respond
+ based on the result, but for now, my immediate response should just be the
+ function call.\n\nIn the previous interaction, after the function call, there
+ was a \"Function:\" response, and then I provided a final answer. I need to
+ follow the same pattern: output the function call, and presumably the system
+ will handle the execution.\n\nMy response should only include the
+ if that''s the next step, which it is.\n\nFinally, keep the response clear
+ and not verbose. So, just the function call.","tool_calls":[{"id":"call_26090001","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":460,"completion_tokens":45,"total_tokens":761,"prompt_tokens_details":{"text_tokens":460,"audio_tokens":0,"image_tokens":0,"cached_tokens":396},"completion_tokens_details":{"reasoning_tokens":256,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 09:57:45 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:58:02 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 09:58:02 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:58:15 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"6ca829ea-cd5a-4b32-323b-767fc49dbe3b_us-east-1","object":"chat.completion","created":1756979882,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking for the weather in Paris with coordinates (48.8575, 2.3514).
+ This matches the \"weather\" function, which requires latitude and longitude.\n\nIn
+ the conversation history, I already called the weather function for this in
+ my previous response, and the system simulated the function response.\n\nNow,
+ in this turn, I need to respond to the user''s new query. The user''s message
+ is: \"What''s the weather in Paris? (48.8575, 2.3514)\"\n\nI should call the
+ function again since it''s a new request.\n\nMy response should include the
+ function call if it''s the logical next step.\n\nAfter the function call,
+ the system will provide the function output, and then I need to formulate
+ a final response.\n\nIn the previous interaction, after my function call,
+ the system output the function result, and then I gave a final response.\n\nSo,
+ for this, I should:\n\n1. Make the function call.\n\n2. Once the function
+ responds, use that to answer the user.\n\nIn the conversation log, it shows:\n\n-
+ I made a function call.\n\n- Then, \"Function: Current weather...\" was output.\n\n-
+ Then, I responded with a summary.\n\nBut in this new turn, I need to do the
+ same.\n\nThe user''s message is clear, so I''ll call the function.\n\nFinal
+ decision: Output the function call in the specified format.","tool_calls":[{"id":"call_58629948","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":614,"completion_tokens":45,"total_tokens":940,"prompt_tokens_details":{"text_tokens":614,"audio_tokens":0,"image_tokens":0,"cached_tokens":532},"completion_tokens_details":{"reasoning_tokens":281,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 09:58:14 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:58:22 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"70353a6f-3043-e731-8a07-ee93fdecf85e_us-east-1","object":"chat.completion","created":1756979895,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking for the weather in Paris with coordinates (48.8575, 2.3514).
+ This matches the available function \"weather\", which requires latitude and
+ longitude.\n\nIn the conversation history, I''ve already handled a similar
+ request for Berlin, and now this is a new one for Paris.\n\nFrom the system
+ prompt, I should call the function if it''s the logical next step. So, I need
+ to trigger the weather function with the provided coordinates.\n\nThe response
+ format should be: If I''m calling a function, use tags with
+ the exact JSON format.\n\nAfter the function call, the system will presumably
+ execute it and provide the result, as seen in previous interactions.\n\nIn
+ the history, after my function call, there''s a \"Function:\" response, and
+ then I respond based on that.\n\nIn my previous response, I directly output
+ the final answer after the function call was simulated.\n\nLooking at the
+ history:\n\n- I output: ...\n\n- Then, there''s
+ a \"Function: Current weather...\" line, which seems to be the system''s response.\n\n-
+ Then, in the next assistant message, I output the final answer.\n\nIn this
+ new interaction, I need to start by making the function call.\n\nThe user''s
+ query is clear: weather for Paris with given coordinates.\n\nSo, my response
+ should be just the function call.\n\nFinally, ensure the JSON is correctly
+ formatted.\n\nAction: \"weather\"\n\nAction input: { \"latitude\": \"48.8575\",
+ \"longitude\": \"2.3514\" }\n\nYes.","tool_calls":[{"id":"call_71779672","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":691,"completion_tokens":45,"total_tokens":1048,"prompt_tokens_details":{"text_tokens":691,"audio_tokens":0,"image_tokens":0,"cached_tokens":609},"completion_tokens_details":{"reasoning_tokens":312,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 09:58:22 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"},{"role":"assistant","content":"","tool_calls":[{"id":"call_71779672","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_71779672"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:58:32 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"936f5f6f-3d03-a7cb-b8a9-d9a015522477_us-east-1","object":"chat.completion","created":1756979903,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking for the weather in Paris with coordinates (48.8575, 2.3514).
+ This matches the available function \"weather\" which requires latitude and
+ longitude.\n\nIn the conversation history, I''ve already handled similar requests
+ by calling the weather function. The system prompt instructs me to call functions
+ when they can advance the user''s request.\n\nThe previous responses show
+ a pattern: I call the function, then the function response is provided, and
+ I respond with a summary.\n\nIn this specific interaction, the assistant has
+ already called the function multiple times in the history, and now I''m responding
+ to the latest user message.\n\nLooking at the history:\n\n- Human asks about
+ Berlin, I call the function.\n\n- Then, Function response is given.\n\n- I
+ summarize it.\n\n- Human asks about Paris, I call the function.\n\n- Function
+ response is given.\n\n- But in the provided text, it seems like there''s a
+ loop or repetition in the history.\n\nIn the actual response I need to give,
+ for this new message, I should treat it as a new request.\n\nThe user''s latest
+ message is: \"What''s the weather in Paris? (48.8575, 2.3514)\"\n\nSo, I need
+ to call the weather function with those parameters.\n\nThe system prompt says:
+ \"If the user''s query can be addressed using an available function, and all
+ required parameters are provided or obviously inferable, trigger a function
+ call...\"\n\nHere, parameters are provided: latitude \"48.8575\" and longitude
+ \"2.3514\".\n\nTherefore, I should make a function call.\n\nMy response should
+ be in the format: {...}\n\nAfter the function
+ call, presumably, the function response will be handled externally, and I
+ might need to summarize it in a follow-up, but for now, my immediate action
+ is to call the function.\n\nIn the history, it shows multiple calls, but that''s
+ just the conversation log. For this turn, I respond to the current human message.\n\nSo,
+ my response should be:\n\n{\n \"action\": \"weather\",\n \"action_input\":
+ {\n \"latitude\": \"48.8575\",\n \"longitude\": \"2.3514\"\n }\n}","tool_calls":[{"id":"call_69075267","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":768,"completion_tokens":45,"total_tokens":1263,"prompt_tokens_details":{"text_tokens":768,"audio_tokens":0,"image_tokens":0,"cached_tokens":686},"completion_tokens_details":{"reasoning_tokens":450,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 09:58:32 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"},{"role":"assistant","content":"","tool_calls":[{"id":"call_71779672","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_71779672"},{"role":"assistant","content":"","tool_calls":[{"id":"call_69075267","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_69075267"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:58:39 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"371030bb-a536-fcb7-c5f4-eb4556f8f430_us-east-1","object":"chat.completion","created":1756979912,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking for the weather in Paris with coordinates (48.8575, 2.3514).
+ This matches the available function \"weather\", which requires latitude and
+ longitude.\n\nIn the conversation history, I''ve already handled similar requests
+ by calling the weather function. Each time, the function call is made, and
+ then the response is provided.\n\nLooking at the system prompt: I need to
+ call functions decisively if the logical next step is to use a provided function.
+ The query provides all required parameters: latitude and longitude.\n\nSo,
+ I should respond with a function call in the specified JSON format within
+ tags.\n\nAfter the function call, in the history, it seems
+ like the function''s response is simulated, and then I provide a final answer.
+ But my role is to only output the function call when appropriate, and then
+ perhaps respond based on the result.\n\nIn the initial setup, after I make
+ a function call, the system provides the function''s output, and then I continue.\n\nFor
+ this turn, the logical step is to call the weather function with the given
+ parameters.\n\nStructure my response:\n- Use tags with the
+ exact format.\n- Action: \"weather\"\n- Action input: { \"latitude\": \"48.8575\",
+ \"longitude\": \"2.3514\" }\n\nFinally, end my response there, as per the
+ instructions.\n\nThe prompt says: \"You are only permitted to use the
+ tag in your response only if you have made the decision to call a function.\"\n\nSo,
+ my response should just be the function call.","tool_calls":[{"id":"call_47468907","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":845,"completion_tokens":45,"total_tokens":1210,"prompt_tokens_details":{"text_tokens":845,"audio_tokens":0,"image_tokens":0,"cached_tokens":3},"completion_tokens_details":{"reasoning_tokens":320,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 09:58:39 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"},{"role":"assistant","content":"","tool_calls":[{"id":"call_71779672","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_71779672"},{"role":"assistant","content":"","tool_calls":[{"id":"call_69075267","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_69075267"},{"role":"assistant","content":"","tool_calls":[{"id":"call_47468907","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_47468907"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:58:46 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"d48d021e-5312-e38e-f9d4-f86d7153081f_us-east-1","object":"chat.completion","created":1756979920,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking for the weather in Paris with coordinates (48.8575, 2.3514).
+ This matches the available function \"weather\" which requires latitude and
+ longitude.\n\nIn the conversation history, I''ve already handled similar requests.
+ For each request, I call the function and then respond based on the result.\n\nLooking
+ at the pattern:\n\n- When I call for weather, the system responds
+ with the weather data.\n\n- Then, in my next response, I should provide the
+ final answer to the user.\n\nIn the history, it seems like there''s a loop,
+ but that''s probably an error in the simulation. In reality, for this new
+ query, I need to:\n\n1. Recognize that this is a new request.\n\n2. Call the
+ weather function with the provided coordinates.\n\nSo, for this turn, the
+ logical step is to call the function again, just like I did for the previous
+ queries.\n\nMy response should be: with the action \"weather\"
+ and the appropriate parameters.\n\nAfter the function call, the system will
+ provide the result, and then I can respond with a user-friendly message.\n\nFinally,
+ keep the response clear and not verbose.\n\nSo, construct the function call:\n\n-
+ Action: \"weather\"\n\n- Action Input: {\"latitude\": \"48.8575\", \"longitude\":
+ \"2.3514\"}\n\nThat matches exactly.","tool_calls":[{"id":"call_34082041","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":922,"completion_tokens":45,"total_tokens":1245,"prompt_tokens_details":{"text_tokens":922,"audio_tokens":0,"image_tokens":0,"cached_tokens":840},"completion_tokens_details":{"reasoning_tokens":278,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 09:58:46 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"},{"role":"assistant","content":"","tool_calls":[{"id":"call_71779672","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_71779672"},{"role":"assistant","content":"","tool_calls":[{"id":"call_69075267","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_69075267"},{"role":"assistant","content":"","tool_calls":[{"id":"call_47468907","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_47468907"},{"role":"assistant","content":"","tool_calls":[{"id":"call_34082041","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_34082041"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:58:55 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 09:58:55 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"},{"role":"assistant","content":"","tool_calls":[{"id":"call_71779672","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_71779672"},{"role":"assistant","content":"","tool_calls":[{"id":"call_69075267","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_69075267"},{"role":"assistant","content":"","tool_calls":[{"id":"call_47468907","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_47468907"},{"role":"assistant","content":"","tool_calls":[{"id":"call_34082041","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_34082041"},{"role":"assistant","content":"","tool_calls":[{"id":"call_96945262","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_96945262"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:59:06 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 09:59:06 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"},{"role":"assistant","content":"","tool_calls":[{"id":"call_71779672","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_71779672"},{"role":"assistant","content":"","tool_calls":[{"id":"call_69075267","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_69075267"},{"role":"assistant","content":"","tool_calls":[{"id":"call_47468907","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_47468907"},{"role":"assistant","content":"","tool_calls":[{"id":"call_34082041","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_34082041"},{"role":"assistant","content":"","tool_calls":[{"id":"call_96945262","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_96945262"},{"role":"assistant","content":"","tool_calls":[{"id":"call_21042674","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_21042674"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:59:28 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 09:59:28 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"},{"role":"assistant","content":"","tool_calls":[{"id":"call_71779672","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_71779672"},{"role":"assistant","content":"","tool_calls":[{"id":"call_69075267","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_69075267"},{"role":"assistant","content":"","tool_calls":[{"id":"call_47468907","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_47468907"},{"role":"assistant","content":"","tool_calls":[{"id":"call_34082041","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_34082041"},{"role":"assistant","content":"","tool_calls":[{"id":"call_96945262","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_96945262"},{"role":"assistant","content":"","tool_calls":[{"id":"call_21042674","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_21042674"},{"role":"assistant","content":"","tool_calls":[{"id":"call_98740275","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_98740275"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:59:39 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 09:59:39 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"},{"role":"assistant","content":"","tool_calls":[{"id":"call_71779672","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_71779672"},{"role":"assistant","content":"","tool_calls":[{"id":"call_69075267","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_69075267"},{"role":"assistant","content":"","tool_calls":[{"id":"call_47468907","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_47468907"},{"role":"assistant","content":"","tool_calls":[{"id":"call_34082041","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_34082041"},{"role":"assistant","content":"","tool_calls":[{"id":"call_96945262","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_96945262"},{"role":"assistant","content":"","tool_calls":[{"id":"call_21042674","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_21042674"},{"role":"assistant","content":"","tool_calls":[{"id":"call_98740275","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_98740275"},{"role":"assistant","content":"","tool_calls":[{"id":"call_79007562","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_79007562"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:59:45 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"b6a2127e-e06e-a1bf-39b1-4b9e2a2f0b1b_us-east-1","object":"chat.completion","created":1756979979,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking for the weather in Paris with coordinates (48.8575, 2.3514).
+ This matches the available function \"weather\" which requires latitude and
+ longitude.\n\nIn the conversation history, I''ve been responding to similar
+ queries by calling the weather function. However, looking at the pattern,
+ it seems like there''s a loop or repetition in the responses, but I need to
+ handle this new query independently.\n\nThe system prompt instructs: \"If
+ the user''s query can be addressed using an available function, and all required
+ parameters are provided or obviously inferable, trigger a function call.\"\n\nThe
+ parameters are provided: latitude \"48.8575\" and longitude \"2.3514\".\n\nSo,
+ I should call the weather function.\n\nAfter the function call, in previous
+ interactions, there''s a \"Function:\" response that provides the weather
+ data, and then I respond with a summary.\n\nIn my response structure, I need
+ to output the function call in the specified JSON format within
+ tags.\n\nFinally, keep the response clear and not verbose.\n\nMy response
+ should be:\n\n\n\n{\n\n \"action\": \"weather\",\n\n \"action_input\":
+ {\n\n \"latitude\": \"48.8575\",\n\n \"longitude\": \"2.3514\"\n\n }\n\n}\n\n\n\nThat''s
+ it for now. The system will handle the function execution.","tool_calls":[{"id":"call_39005127","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":1307,"completion_tokens":45,"total_tokens":1626,"prompt_tokens_details":{"text_tokens":1307,"audio_tokens":0,"image_tokens":0,"cached_tokens":1225},"completion_tokens_details":{"reasoning_tokens":274,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 09:59:45 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_84020122","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84020122"},{"role":"assistant","content":"The
+ current weather in Berlin (at latitude 52.5200, longitude 13.4050) is 15°C
+ with wind at 10 km/h."},{"role":"user","content":"What''s the weather in Paris?
+ (48.8575, 2.3514)"},{"role":"assistant","content":"","tool_calls":[{"id":"call_26090001","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_26090001"},{"role":"assistant","content":"","tool_calls":[{"id":"call_16030494","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_16030494"},{"role":"assistant","content":"","tool_calls":[{"id":"call_58629948","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_58629948"},{"role":"assistant","content":"","tool_calls":[{"id":"call_71779672","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_71779672"},{"role":"assistant","content":"","tool_calls":[{"id":"call_69075267","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_69075267"},{"role":"assistant","content":"","tool_calls":[{"id":"call_47468907","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_47468907"},{"role":"assistant","content":"","tool_calls":[{"id":"call_34082041","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_34082041"},{"role":"assistant","content":"","tool_calls":[{"id":"call_96945262","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_96945262"},{"role":"assistant","content":"","tool_calls":[{"id":"call_21042674","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_21042674"},{"role":"assistant","content":"","tool_calls":[{"id":"call_98740275","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_98740275"},{"role":"assistant","content":"","tool_calls":[{"id":"call_79007562","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_79007562"},{"role":"assistant","content":"","tool_calls":[{"id":"call_39005127","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_39005127"}],"stream":false,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:59:57 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 09:59:57 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_with_multi-turn_streaming_conversations.yml b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_with_multi-turn_streaming_conversations.yml
new file mode 100644
index 000000000..65b14741f
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_with_multi-turn_streaming_conversations.yml
@@ -0,0 +1,2099 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"}],"stream":true,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}],"stream_options":{"include_usage":true}}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:30 GMT
+ Content-Type:
+ - text/event-stream
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ Cf-Ray:
+ - ""
+ Cache-Control:
+ - no-cache
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ body:
+ encoding: UTF-8
+ string: |+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"First","role":"assistant"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" asking"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" in"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Berlin"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" has"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" provided"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" coordinates"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" ("}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"52"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"520"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"0"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"13"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"405"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"0"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":").\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"The"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" available"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" which"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" requires"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" two"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" parameters"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"latitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980151,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"longitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"The"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" has"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" given"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" exactly"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" that"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" latitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" as"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"52"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"520"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"0"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" longitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" as"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"13"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"405"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980152,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"0"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"This"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" matches"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"'s"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" requirements"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" perfectly"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" can"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" these"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" values"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" need"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" format"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" as"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" specified"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980154,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" within"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" <"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":">"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" tags"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" in"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" JSON"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" format"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"The"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" exact"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" format"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" <"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":">{\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"action"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_name"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"action"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_input"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" {\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"parameter"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"argument"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"}}\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"So"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" this"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" it"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" be"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" <"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":">{\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"action"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"action"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_input"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" {\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"latitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"52"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"520"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"0"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"longitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"13"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"405"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"0"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"}}\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"After"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" calling"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" shouldn't"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" add"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" any"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" extra"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" text"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" unless"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" necessary"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" as"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" per"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" instructions"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" The"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" system"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" prompt"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" says"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Keep"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" your"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" clear"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":";"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" please"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" do"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" not"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" make"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" your"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" verbose"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"!\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" So"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" if"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I'm"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" just"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" calling"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" that"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" be"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" my"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" entire"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"This"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" single"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-turn"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" action"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" since"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" all"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" info"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" provided"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Once"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" this"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" it"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" handle"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" request"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"I'm"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" ready"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" make"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980155,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980156,"model":"grok-3-mini","choices":[{"index":0,"delta":{"tool_calls":[{"id":"call_84692337","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"},"index":0,"type":"function"}]}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980156,"model":"grok-3-mini","choices":[{"index":0,"delta":{},"finish_reason":"tool_calls"}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"877bfa44-7ba2-f6ad-daaa-910389071a7b_us-east-1","object":"chat.completion.chunk","created":1756980156,"model":"grok-3-mini","choices":[],"usage":{"prompt_tokens":324,"completion_tokens":37,"total_tokens":633,"prompt_tokens_details":{"text_tokens":324,"audio_tokens":0,"image_tokens":0,"cached_tokens":323},"completion_tokens_details":{"reasoning_tokens":272,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}
+
+ data: [DONE]
+
+ recorded_at: Thu, 04 Sep 2025 10:02:36 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","tool_calls":[{"id":"call_84692337","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84692337"}],"stream":true,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}],"stream_options":{"include_usage":true}}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:37 GMT
+ Content-Type:
+ - text/event-stream
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ Cf-Ray:
+ - ""
+ Cache-Control:
+ - no-cache
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 10:02:41 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","tool_calls":[{"id":"call_84692337","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84692337"},{"role":"assistant","content":"The
+ current weather in Berlin (52.5200, 13.4050) is 15°C with wind at 10 km/h."},{"role":"user","content":"What''s
+ the weather in Paris? (48.8575, 2.3514)"}],"stream":true,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}],"stream_options":{"include_usage":true}}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:41 GMT
+ Content-Type:
+ - text/event-stream
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ Cf-Ray:
+ - ""
+ Cache-Control:
+ - no-cache
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Report-To:
+ - '{"endpoints":[{"url":"https:\/\/csp-reporting.cloudflare.com\/cdn-cgi\/script_monitor\/report?m=Jl94hgCCLeAK8LhZnxtaEL4xoA0PyHAtdaMjo_EKNJo-1756980161-1.0.1.1-7hIa50GWRA5nixWB0AwYwzDQK0huEV_dbiNmtIGpt_i9.nSMGXLvEgpbbGh4zNY6caD3hMxmt_NPiHa5dHfB4Wdm9750ILJkUyZJAbP73Ek8ixp1bKJR_DhRz6SKzBKTr7JDVuPmNcfNALyOoMxG6K5bTytu8KA.kveMXR9XcMM"}],"group":"cf-csp-endpoint","max_age":86400}'
+ Content-Security-Policy-Report-Only:
+ - script-src 'none'; report-uri https://csp-reporting.cloudflare.com/cdn-cgi/script_monitor/report?m=Jl94hgCCLeAK8LhZnxtaEL4xoA0PyHAtdaMjo_EKNJo-1756980161-1.0.1.1-7hIa50GWRA5nixWB0AwYwzDQK0huEV_dbiNmtIGpt_i9.nSMGXLvEgpbbGh4zNY6caD3hMxmt_NPiHa5dHfB4Wdm9750ILJkUyZJAbP73Ek8ixp1bKJR_DhRz6SKzBKTr7JDVuPmNcfNALyOoMxG6K5bTytu8KA.kveMXR9XcMM;
+ report-to cf-csp-endpoint
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ body:
+ encoding: UTF-8
+ string: |+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"First","role":"assistant"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" asking"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" in"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Paris"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" providing"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" coordinates"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" ("}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"48"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"857"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"5"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"351"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"4"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":").\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" have"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" available"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" which"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" gets"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" current"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" location"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" using"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" latitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" longitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980162,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"The"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" required"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" parameters"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" are"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"latitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"longitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" which"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" are"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" both"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" provided"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" in"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user's"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" query"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"This"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" matches"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" directly"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" available"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" so"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" it"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"In"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" previous"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" interaction"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" called"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" then"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" responded"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980163,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" result"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Now"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" need"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" do"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" same"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"The"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" system"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" prompt"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" says"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" If"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user's"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" query"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" can"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" be"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" addressed"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" using"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" an"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" available"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" all"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" required"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" parameters"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" are"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" provided"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" trigger"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"So"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" output"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" <"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":">"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" tag"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" appropriate"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" action"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980164,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" parameters"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"After"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" assume"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" system"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" will"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" provide"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" result"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" then"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" can"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" respond"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" based"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" on"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" that"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"In"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" conversation"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" history"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" after"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" my"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" there's"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" then"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" gave"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" final"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" answer"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"For"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" this"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" just"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" need"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980165,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" make"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Structure"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" my"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Only"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" include"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" <"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":">"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" tag"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" if"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I'm"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" calling"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Final"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" decision"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" given"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" latitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" longitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980166,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980167,"model":"grok-3-mini","choices":[{"index":0,"delta":{"tool_calls":[{"id":"call_08386256","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"},"index":0,"type":"function"}]}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980167,"model":"grok-3-mini","choices":[{"index":0,"delta":{},"finish_reason":"tool_calls"}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"6f42c364-450f-7f44-77a8-b77a2604150a_us-east-1","object":"chat.completion.chunk","created":1756980167,"model":"grok-3-mini","choices":[],"usage":{"prompt_tokens":456,"completion_tokens":45,"total_tokens":756,"prompt_tokens_details":{"text_tokens":456,"audio_tokens":0,"image_tokens":0,"cached_tokens":406},"completion_tokens_details":{"reasoning_tokens":255,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}
+
+ data: [DONE]
+
+ recorded_at: Thu, 04 Sep 2025 10:02:47 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","tool_calls":[{"id":"call_84692337","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84692337"},{"role":"assistant","content":"The
+ current weather in Berlin (52.5200, 13.4050) is 15°C with wind at 10 km/h."},{"role":"user","content":"What''s
+ the weather in Paris? (48.8575, 2.3514)"},{"role":"assistant","tool_calls":[{"id":"call_08386256","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_08386256"}],"stream":true,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}],"stream_options":{"include_usage":true}}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:47 GMT
+ Content-Type:
+ - text/event-stream
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ Cf-Ray:
+ - ""
+ Cache-Control:
+ - no-cache
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ body:
+ encoding: UTF-8
+ string: |+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980167,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"First","role":"assistant"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" asking"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" in"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Paris"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" providing"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" coordinates"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" ("}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"48"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"857"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"5"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"351"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"4"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":")."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" This"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" matches"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" available"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" which"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" requires"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" latitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" longitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" need"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" provided"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" parameters"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" So"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980168,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" use"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" <"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":">"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" tag"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" action"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" action"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_input"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" containing"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" latitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" longitude"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"After"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" calling"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" system"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" will"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" presumably"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" respond"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" data"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" as"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" it"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" did"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" in"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" previous"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" interaction"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"In"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" previous"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" after"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" my"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" there"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980169,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" was"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" then"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" summarized"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" it"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" follow"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" same"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" pattern"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" make"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" then"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" in"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" subsequent"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" provide"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" information"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" based"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" on"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"'s"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" output"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Looking"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" at"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" conversation"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" history"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" made"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980170,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Then"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" there's"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" data"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Finally"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" responded"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-friendly"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" message"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"In"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" this"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" setup"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" need"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" output"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980171,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" first"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"The"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" system"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" prompt"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" says"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"You"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" can"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" use"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" multiple"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980172,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" functions"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" or"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" multiple"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" times"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" same"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" this"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" use"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" multiple"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" <"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":">"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" tags"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" But"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" in"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" this"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" case"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" it's"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" single"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"My"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" only"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" include"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" <"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"_call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":">"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" if"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I'm"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" deciding"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" which"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" am"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"So"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" this"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" query"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" my"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" immediate"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" be"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"After"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" that"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" assuming"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" executes"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I'll"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" get"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" result"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" can"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" respond"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" accordingly"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"In"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" example"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" provided"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" after"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" my"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" next"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" part"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Current"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"...\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" then"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" my"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" final"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"In"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" my"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" output"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" just"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" need"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" provide"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" now"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Final"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980173,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" decision"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Call"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" weather"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" function"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" given"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" coordinates"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980174,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980175,"model":"grok-3-mini","choices":[{"index":0,"delta":{"tool_calls":[{"id":"call_56232890","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"},"index":0,"type":"function"}]}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980175,"model":"grok-3-mini","choices":[{"index":0,"delta":{},"finish_reason":"tool_calls"}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"f7878e9e-8209-9742-58e2-3654346d4be9_us-east-1","object":"chat.completion.chunk","created":1756980175,"model":"grok-3-mini","choices":[],"usage":{"prompt_tokens":533,"completion_tokens":45,"total_tokens":927,"prompt_tokens_details":{"text_tokens":533,"audio_tokens":0,"image_tokens":0,"cached_tokens":451},"completion_tokens_details":{"reasoning_tokens":349,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}
+
+ data: [DONE]
+
+ recorded_at: Thu, 04 Sep 2025 10:02:55 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the weather in Berlin? (52.5200, 13.4050)"},{"role":"assistant","tool_calls":[{"id":"call_84692337","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"52.5200\",\"longitude\":\"13.4050\"}"}}]},{"role":"tool","content":"Current
+ weather at 52.5200, 13.4050: 15°C, Wind: 10 km/h","tool_call_id":"call_84692337"},{"role":"assistant","content":"The
+ current weather in Berlin (52.5200, 13.4050) is 15°C with wind at 10 km/h."},{"role":"user","content":"What''s
+ the weather in Paris? (48.8575, 2.3514)"},{"role":"assistant","tool_calls":[{"id":"call_08386256","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_08386256"},{"role":"assistant","tool_calls":[{"id":"call_56232890","type":"function","function":{"name":"weather","arguments":"{\"latitude\":\"48.8575\",\"longitude\":\"2.3514\"}"}}]},{"role":"tool","content":"Current
+ weather at 48.8575, 2.3514: 15°C, Wind: 10 km/h","tool_call_id":"call_56232890"}],"stream":true,"tools":[{"type":"function","function":{"name":"weather","description":"Gets
+ current weather for a location","parameters":{"type":"object","properties":{"latitude":{"type":"string","description":"Latitude
+ (e.g., 52.5200)"},"longitude":{"type":"string","description":"Longitude (e.g.,
+ 13.4050)"}},"required":["latitude","longitude"]}}}],"stream_options":{"include_usage":true}}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:55 GMT
+ Content-Type:
+ - text/event-stream
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ Cf-Ray:
+ - ""
+ Cache-Control:
+ - no-cache
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 10:03:06 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_without_parameters.yml b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_without_parameters.yml
new file mode 100644
index 000000000..e80fd5da4
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_function_calling_xai_grok-3-mini_can_use_tools_without_parameters.yml
@@ -0,0 +1,151 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the best language to learn?"}],"stream":false,"tools":[{"type":"function","function":{"name":"best_language_to_learn","description":"Gets
+ the best language to learn","parameters":{"type":"object","properties":{},"required":[]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:23 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"a14b22fd-e10c-16c7-cccf-f6b09327d39d_us-east-1","object":"chat.completion","created":1756980137,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"","reasoning_content":"First,
+ the user is asking, \"What''s the best language to learn?\" This could mean
+ a programming language, a natural language, or something else. But looking
+ at the available functions, there''s only one: \"best_language_to_learn\".
+ Its description is \"Gets the best language to learn\", and it has no required
+ parameters.\n\nThe function''s name and description suggest it''s designed
+ exactly for this query. Since no parameters are needed, I can call this function
+ directly.\n\nI need to remember the format: If I decide to call a function,
+ I use tags with JSON inside, like {\"action\":
+ \"function_name\", \"action_input\": {}}.\n\nIn this case,
+ the action is \"best_language_to_learn\", and action_input is an empty object
+ because there are no parameters.\n\nAfter calling the function, I should wait
+ for the response, but in this setup, I''m just supposed to output the function
+ call if that''s the next step.\n\nThe instructions say: \"Keep your response
+ to user clear; please do not make your response verbose!\" So, I shouldn''t
+ add extra explanations unless necessary.\n\nThis seems straightforward. No
+ need for clarification since the function matches the query perfectly.\n\nFinally,
+ output only the function call in the specified format.","tool_calls":[{"id":"call_06589537","function":{"name":"best_language_to_learn","arguments":"{}"},"type":"function"}],"refusal":null},"finish_reason":"tool_calls"}],"usage":{"prompt_tokens":273,"completion_tokens":24,"total_tokens":552,"prompt_tokens_details":{"text_tokens":273,"audio_tokens":0,"image_tokens":0,"cached_tokens":231},"completion_tokens_details":{"reasoning_tokens":255,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 10:02:23 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"What''s
+ the best language to learn?"},{"role":"assistant","content":"","tool_calls":[{"id":"call_06589537","type":"function","function":{"name":"best_language_to_learn","arguments":"{}"}}]},{"role":"tool","content":"Ruby","tool_call_id":"call_06589537"}],"stream":false,"tools":[{"type":"function","function":{"name":"best_language_to_learn","description":"Gets
+ the best language to learn","parameters":{"type":"object","properties":{},"required":[]}}}]}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:29 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"39f37419-ee26-427d-a545-bc42cc2a5148_us-east-1","object":"chat.completion","created":1756980143,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"The
+ best language to learn is Ruby.","reasoning_content":"First, the user asked:
+ \"What''s the best language to learn?\" This directly matches the available
+ function \"best_language_to_learn\".\n\nThe function''s description is \"Gets
+ the best language to learn\", and it has no required parameters, so I can
+ call it directly.\n\nIn my previous response, I already called the function,
+ and now I''m receiving a response from it: \"Function: Ruby\". This means
+ the function call has been executed, and it''s providing the result.\n\nI
+ need to handle this in the context of the conversation. The system prompt
+ says: \"In multi-turn conversations, continue calling relevant functions to
+ advance the user''s request until either the user''s request is fully fulfilled
+ or you need more information to call a function.\"\n\nHowever, I''ve already
+ called the function, and now I have the result. So, I should use this result
+ to respond to the user.\n\nThe prompt also says: \"Keep your response to user
+ clear; please do not make your response verbose!\" So, I need to be concise.\n\nMy
+ role is to respond based on the function''s output. Since the function returned
+ \"Ruby\", I can now inform the user of that.\n\nI shouldn''t call another
+ function unless it''s needed. The user''s request seems fulfilled with this
+ information.\n\nFinal response: I should directly tell the user the result
+ from the function call.","refusal":null},"finish_reason":"stop"}],"usage":{"prompt_tokens":308,"completion_tokens":8,"total_tokens":579,"prompt_tokens_details":{"text_tokens":308,"audio_tokens":0,"image_tokens":0,"cached_tokens":3},"completion_tokens_details":{"reasoning_tokens":263,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 10:02:29 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_real_error_scenarios_xai_grok-3-mini_handles_context_length_exceeded_errors.yml b/spec/fixtures/vcr_cassettes/chat_real_error_scenarios_xai_grok-3-mini_handles_context_length_exceeded_errors.yml
new file mode 100644
index 000000000..881ac1614
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_real_error_scenarios_xai_grok-3-mini_handles_context_length_exceeded_errors.yml
@@ -0,0 +1,54 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":""},{"role":"assistant","content":""},{"role":"user","content":""},{"role":"assistant","content":""},{"role":"user","content":""},{"role":"assistant","content":""},{"role":"user","content":""},{"role":"assistant","content":""},{"role":"user","content":""},{"role":"assistant","content":""},{"role":"user","content":"Hi"}],"stream":false}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 400
+ message: Bad Request
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:00:20 GMT
+ Content-Type:
+ - application/json
+ Content-Length:
+ - '142'
+ Connection:
+ - keep-alive
+ Cf-Ray:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ body:
+ encoding: UTF-8
+ string: '{"code":"Client specified an invalid argument","error":"This model''s
+ maximum prompt length is 131072 but the request contains 625078 tokens."}'
+ recorded_at: Thu, 04 Sep 2025 10:00:20 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_streaming_responses_xai_grok-3-mini_reports_consistent_token_counts_compared_to_non-streaming.yml b/spec/fixtures/vcr_cassettes/chat_streaming_responses_xai_grok-3-mini_reports_consistent_token_counts_compared_to_non-streaming.yml
new file mode 100644
index 000000000..d4c2a1b0f
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_streaming_responses_xai_grok-3-mini_reports_consistent_token_counts_compared_to_non-streaming.yml
@@ -0,0 +1,504 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"Count
+ from 1 to 3, respond with a comma-delimited list of integers with no spaces."}],"stream":true,"temperature":0.0,"stream_options":{"include_usage":true}}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:01:56 GMT
+ Content-Type:
+ - text/event-stream
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ Cf-Ray:
+ - ""
+ Cache-Control:
+ - no-cache
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Report-To:
+ - '{"endpoints":[{"url":"https:\/\/csp-reporting.cloudflare.com\/cdn-cgi\/script_monitor\/report?m=5MqbSYvRE6SYbinAu0mhe3_fCTFbwa8e4fHt2p0SVHs-1756980116-1.0.1.1-bixxm7ipeOHmVvu8sAG7GWs0.oH3UB207CRk.866TKLukq4dNq0y3ucegFTM5ghfKHlYilj8.twLWkw8FeEbnworXRuHuXVpPypa_UpVzy.d_gRxAf5QRSPDUiN.Q2EaYawrDOGORMc6AuMTPERvMkiqGZQNzEyHUEg.aYo3aww"}],"group":"cf-csp-endpoint","max_age":86400}'
+ Content-Security-Policy-Report-Only:
+ - script-src 'none'; connect-src 'none'; report-uri https://csp-reporting.cloudflare.com/cdn-cgi/script_monitor/report?m=5MqbSYvRE6SYbinAu0mhe3_fCTFbwa8e4fHt2p0SVHs-1756980116-1.0.1.1-bixxm7ipeOHmVvu8sAG7GWs0.oH3UB207CRk.866TKLukq4dNq0y3ucegFTM5ghfKHlYilj8.twLWkw8FeEbnworXRuHuXVpPypa_UpVzy.d_gRxAf5QRSPDUiN.Q2EaYawrDOGORMc6AuMTPERvMkiqGZQNzEyHUEg.aYo3aww;
+ report-to cf-csp-endpoint
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ body:
+ encoding: UTF-8
+ string: |+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"First","role":"assistant"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" asked"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" me"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Count"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" from"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" respond"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" comma"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-del"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"imited"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" list"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" of"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" integers"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" no"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" spaces"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\"\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"So"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" counting"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" from"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" means"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" numbers"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980117,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"They"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" want"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" comma"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-del"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"imited"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" list"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" which"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" means"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" separating"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" numbers"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" commas"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" like"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"And"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" no"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" spaces"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" means"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" no"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" spaces"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" after"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" commas"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" so"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" not"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980118,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" but"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"My"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" be"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" exactly"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" that"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" list"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" of"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" integers"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" from"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" separated"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" by"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" commas"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" no"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" spaces"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"As"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" an"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" AI"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" respond"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" directly"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" accurately"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" without"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" adding"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" extra"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" text"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" unless"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" specified"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980119,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" The"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" instruction"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"respond"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" with"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" that"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" list"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" so"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I'll"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" just"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" output"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" list"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Final"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"delta":{},"finish_reason":"stop"}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"d506ab64-09e1-a241-c1fc-366b471dac95_us-east-1","object":"chat.completion.chunk","created":1756980120,"model":"grok-3-mini","choices":[],"usage":{"prompt_tokens":28,"completion_tokens":5,"total_tokens":211,"prompt_tokens_details":{"text_tokens":28,"audio_tokens":0,"image_tokens":0,"cached_tokens":2},"completion_tokens_details":{"reasoning_tokens":178,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}
+
+ data: [DONE]
+
+ recorded_at: Thu, 04 Sep 2025 10:02:00 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"Count
+ from 1 to 3, respond with a comma-delimited list of integers with no spaces."}],"stream":false,"temperature":0.0}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:02:04 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"9fae27d4-7431-cea7-f56c-5e4756b55823_us-east-1","object":"chat.completion","created":1756980120,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"1,2,3","reasoning_content":"First,
+ the user asked me to \"Count from 1 to 3, respond with a comma-delimited list
+ of integers with no spaces.\"\n\nSo, counting from 1 to 3 means the numbers:
+ 1, 2, and 3.\n\nThey want a comma-delimited list, which means separating the
+ numbers with commas, like 1,2,3.\n\nAnd they specified \"with no spaces,\"
+ so no spaces after the commas or anywhere else. So, it should be exactly:
+ 1,2,3\n\nLet me double-check the instructions: \"respond with a comma-delimited
+ list of integers with no spaces.\" Yes, that''s clear.\n\nMy response should
+ be just that string: \"1,2,3\"\n\nI need to make sure I''m not adding any
+ extra text. The user said \"respond with,\" so probably just the list itself.\n\nAs
+ an AI, I should be helpful and follow instructions precisely.\n\nFinal response:
+ 1,2,3","refusal":null},"finish_reason":"stop"}],"usage":{"prompt_tokens":28,"completion_tokens":5,"total_tokens":230,"prompt_tokens_details":{"text_tokens":28,"audio_tokens":0,"image_tokens":0,"cached_tokens":2},"completion_tokens_details":{"reasoning_tokens":197,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 10:02:04 GMT
+recorded_with: VCR 6.3.1
+...
diff --git a/spec/fixtures/vcr_cassettes/chat_streaming_responses_xai_grok-3-mini_supports_streaming_responses.yml b/spec/fixtures/vcr_cassettes/chat_streaming_responses_xai_grok-3-mini_supports_streaming_responses.yml
new file mode 100644
index 000000000..482e90f0c
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_streaming_responses_xai_grok-3-mini_supports_streaming_responses.yml
@@ -0,0 +1,573 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":"Count
+ from 1 to 3"}],"stream":true,"stream_options":{"include_usage":true}}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:01:51 GMT
+ Content-Type:
+ - text/event-stream
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ Cf-Ray:
+ - ""
+ Cache-Control:
+ - no-cache
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ body:
+ encoding: UTF-8
+ string: |+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"First","role":"assistant"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" user"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" said"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Count"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" from"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980111,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" This"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" seems"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" straightforward"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" They"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" want"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" me"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" count"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" numbers"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" from"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"As"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" an"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" AI"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" respond"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" help"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"fully"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" accurately"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" The"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" simplest"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" way"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" list"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" numbers"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" could"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" make"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" it"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" more"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980112,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" engaging"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" For"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" example"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" say"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Sure"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" let's"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" count"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" from"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"!\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" But"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" keep"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" it"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" simple"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" unless"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" specified"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" otherwise"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" there"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" any"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" context"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" or"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" additional"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" instructions"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"?"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" This"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" is"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" standalone"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" message"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" so"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" no"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" prior"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" context"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980113,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Possible"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" interpretations"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Literal"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" counting"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Just"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" say"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" numbers"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Could"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" be"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" test"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" for"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" basic"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" functionality"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"-"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Maybe"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" it's"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" part"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" of"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" a"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" larger"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" interaction"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" but"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" I"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" don't"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" have"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" that"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" info"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"My"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" should"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" be"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" clear"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" concise"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" and"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" polite"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Structure"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" A"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980114,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"cknowledge"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" request"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Okay"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":",\""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" or"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Sure"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\"\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" Perform"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980115,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" action"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" State"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" numbers"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" End"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" politely"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" if"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" needed"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Final"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" response"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" idea"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" \""}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"Counting"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" from"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":":"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":".\"\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"That"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" seems"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":" perfect"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"reasoning_content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"Sure"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":" here's"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":" the"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":" count"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":" from"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":" to"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":":\n\n"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"1"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"2"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":","}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":" "}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"3"}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{"content":"."}}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[{"index":0,"delta":{},"finish_reason":"stop"}],"system_fingerprint":"fp_6a09108ff5"}
+
+ data: {"id":"cc44c127-3603-b8a4-b1a1-477148213556_us-east-1","object":"chat.completion.chunk","created":1756980116,"model":"grok-3-mini","choices":[],"usage":{"prompt_tokens":14,"completion_tokens":20,"total_tokens":267,"prompt_tokens_details":{"text_tokens":14,"audio_tokens":0,"image_tokens":0,"cached_tokens":2},"completion_tokens_details":{"reasoning_tokens":233,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}
+
+ data: [DONE]
+
+ recorded_at: Thu, 04 Sep 2025 10:01:56 GMT
+recorded_with: VCR 6.3.1
+...
diff --git a/spec/fixtures/vcr_cassettes/chat_text_models_xai_grok-3-mini_can_understand_remote_text.yml b/spec/fixtures/vcr_cassettes/chat_text_models_xai_grok-3-mini_can_understand_remote_text.yml
new file mode 100644
index 000000000..b3b2defc9
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_text_models_xai_grok-3-mini_can_understand_remote_text.yml
@@ -0,0 +1,243 @@
+---
+http_interactions:
+- request:
+ method: get
+ uri: https://www.ruby-lang.org/en/about/license.txt
+ body:
+ encoding: US-ASCII
+ string: ''
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Connection:
+ - keep-alive
+ Content-Length:
+ - '2502'
+ Server:
+ - GitHub.com
+ Content-Type:
+ - text/plain; charset=utf-8
+ X-Origin-Cache:
+ - HIT
+ Last-Modified:
+ - Mon, 18 Aug 2025 06:05:44 GMT
+ Access-Control-Allow-Origin:
+ - "*"
+ Etag:
+ - W/"68a2c2b8-9c6"
+ Expires:
+ - Thu, 04 Sep 2025 10:04:52 GMT
+ Cache-Control:
+ - max-age=600
+ X-Proxy-Cache:
+ - MISS
+ X-Github-Request-Id:
+ - 8363:34D603:36C07D:388D3E:68B961EC
+ Accept-Ranges:
+ - bytes
+ Age:
+ - '0'
+ Date:
+ - Thu, 04 Sep 2025 09:59:53 GMT
+ Via:
+ - 1.1 varnish
+ X-Served-By:
+ - cache-lhr-egll1980082-LHR
+ X-Cache:
+ - HIT
+ X-Cache-Hits:
+ - '0'
+ X-Timer:
+ - S1756979993.247111,VS0,VE84
+ Vary:
+ - Accept-Encoding
+ X-Fastly-Request-Id:
+ - 636a59d1252e94d6a4d0b5ab6f48d5ec50faa6b4
+ body:
+ encoding: ASCII-8BIT
+ string: "Ruby is copyrighted free software by Yukihiro Matsumoto .\nYou
+ can redistribute it and/or modify it under either the terms of the\n2-clause
+ BSDL (see the file BSDL), or the conditions below:\n\n 1. You may make and
+ give away verbatim copies of the source form of the\n software without
+ restriction, provided that you duplicate all of the\n original copyright
+ notices and associated disclaimers.\n\n 2. You may modify your copy of the
+ software in any way, provided that\n you do at least ONE of the following:\n\n
+ \ a) place your modifications in the Public Domain or otherwise\n make
+ them Freely Available, such as by posting said\n\t modifications to Usenet
+ or an equivalent medium, or by allowing\n\t the author to include your modifications
+ in the software.\n\n b) use the modified software only within your corporation
+ or\n organization.\n\n c) give non-standard binaries non-standard
+ names, with\n instructions on where to get the original software
+ distribution.\n\n d) make other distribution arrangements with the author.\n\n
+ \ 3. You may distribute the software in object code or binary form,\n provided
+ that you do at least ONE of the following:\n\n a) distribute the binaries
+ and library files of the software,\n\t together with instructions (in the
+ manual page or equivalent)\n\t on where to get the original distribution.\n\n
+ \ b) accompany the distribution with the machine-readable source of\n\t
+ \ the software.\n\n c) give non-standard binaries non-standard names,
+ with\n instructions on where to get the original software distribution.\n\n
+ \ d) make other distribution arrangements with the author.\n\n 4. You
+ may modify and include the part of the software into any other\n software
+ (possibly commercial). But some files in the distribution\n are not written
+ by the author, so that they are not under these terms.\n\n For the list
+ of those files and their copying conditions, see the\n file LEGAL.\n\n
+ \ 5. The scripts and library files supplied as input to or produced as\n output
+ from the software do not automatically fall under the\n copyright of the
+ software, but belong to whomever generated them,\n and may be sold commercially,
+ and may be aggregated with this\n software.\n\n 6. THIS SOFTWARE IS PROVIDED
+ \"AS IS\" AND WITHOUT ANY EXPRESS OR\n IMPLIED WARRANTIES, INCLUDING,
+ WITHOUT LIMITATION, THE IMPLIED\n WARRANTIES OF MERCHANTABILITY AND FITNESS
+ FOR A PARTICULAR\n PURPOSE.\n"
+ recorded_at: Thu, 04 Sep 2025 09:59:53 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":[{"type":"text","text":"What''s
+ in this file?"},{"type":"text","text":"Ruby
+ is copyrighted free software by Yukihiro Matsumoto .\nYou
+ can redistribute it and/or modify it under either the terms of the\n2-clause
+ BSDL (see the file BSDL), or the conditions below:\n\n 1. You may make and
+ give away verbatim copies of the source form of the\n software without
+ restriction, provided that you duplicate all of the\n original copyright
+ notices and associated disclaimers.\n\n 2. You may modify your copy of the
+ software in any way, provided that\n you do at least ONE of the following:\n\n a)
+ place your modifications in the Public Domain or otherwise\n make
+ them Freely Available, such as by posting said\n\t modifications to Usenet
+ or an equivalent medium, or by allowing\n\t the author to include your modifications
+ in the software.\n\n b) use the modified software only within your corporation
+ or\n organization.\n\n c) give non-standard binaries non-standard
+ names, with\n instructions on where to get the original software
+ distribution.\n\n d) make other distribution arrangements with the author.\n\n 3.
+ You may distribute the software in object code or binary form,\n provided
+ that you do at least ONE of the following:\n\n a) distribute the binaries
+ and library files of the software,\n\t together with instructions (in the
+ manual page or equivalent)\n\t on where to get the original distribution.\n\n b)
+ accompany the distribution with the machine-readable source of\n\t the software.\n\n c)
+ give non-standard binaries non-standard names, with\n instructions
+ on where to get the original software distribution.\n\n d) make other
+ distribution arrangements with the author.\n\n 4. You may modify and include
+ the part of the software into any other\n software (possibly commercial). But
+ some files in the distribution\n are not written by the author, so that
+ they are not under these terms.\n\n For the list of those files and their
+ copying conditions, see the\n file LEGAL.\n\n 5. The scripts and library
+ files supplied as input to or produced as\n output from the software do
+ not automatically fall under the\n copyright of the software, but belong
+ to whomever generated them,\n and may be sold commercially, and may be
+ aggregated with this\n software.\n\n 6. THIS SOFTWARE IS PROVIDED \"AS
+ IS\" AND WITHOUT ANY EXPRESS OR\n IMPLIED WARRANTIES, INCLUDING, WITHOUT
+ LIMITATION, THE IMPLIED\n WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ A PARTICULAR\n PURPOSE.\n"}]}],"stream":false}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 10:00:09 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"c466188c-a3d1-66cc-e0f7-fbc0c054b4eb_us-east-1","object":"chat.completion","created":1756979993,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"This
+ file, named `license.txt` and of type `text/plain`, contains the license agreement
+ for the Ruby programming language, authored by Yukihiro Matsumoto. It outlines
+ the terms under which Ruby can be redistributed, modified, and used. Here''s
+ a concise summary of its key points:\n\n### Main Provisions:\n- **Dual Licensing**:
+ Ruby can be used under either the 2-clause BSD License (BSDL, referenced in
+ a separate file) or the specific conditions detailed in this file.\n- **Redistribution
+ of Source Code**:\n - You can freely make and distribute exact copies of
+ the source code, as long as all original copyright notices and disclaimers
+ are included.\n- **Modification**:\n - You may modify the software, but you
+ must do at least one of the following: make the changes publicly available
+ (e.g., via Usenet or similar), limit use to your organization, use non-standard
+ names for modified versions with instructions to the original, or arrange
+ distribution with the author.\n- **Distribution in Binary Form**:\n - You
+ can distribute binaries, but you must include instructions on obtaining the
+ original source, provide the source code with the binaries, use non-standard
+ names with instructions, or make other arrangements with the author.\n- **Incorporation
+ into Other Software**:\n - Parts of Ruby can be modified and included in
+ other programs (even commercial ones), but some files in the distribution
+ have separate licensing terms (detailed in a file called `LEGAL`).\n- **Generated
+ Files**:\n - Scripts or library files created using Ruby are not automatically
+ covered by this license; they belong to their creator and can be sold or distributed
+ independently.\n- **Disclaimer**:\n - The software is provided \"as is,\"
+ with no warranties, including for merchantability or fitness for a particular
+ purpose.\n\nThis license promotes free use while protecting the original copyright.
+ If you''re working with Ruby, it''s advisable to review the full text or consult
+ legal advice for specific use cases.","reasoning_content":"First, the user
+ is asking, \"What''s in this file?\" and they''ve provided the content of
+ a file named ''license.txt'' with a mime_type of ''text/plain''. So, I need
+ to summarize or describe what''s in that file.\n\nThe file appears to be a
+ license for Ruby, written by Yukihiro Matsumoto. It''s a custom license that
+ allows redistribution and modification under certain conditions.\n\nKey points
+ from the content:\n\n- Ruby is copyrighted free software by Yukihiro Matsumoto.\n\n-
+ You can redistribute and modify it under the 2-clause BSDL (which I believe
+ is the 2-clause BSD License) or the conditions listed below.\n\n- The conditions
+ are numbered 1 through 6:\n\n 1. Making and giving away verbatim copies without
+ restriction, as long as all original copyright notices and disclaimers are
+ duplicated.\n\n 2. Modifying the software, but you must do at least one of:
+ make modifications public domain, use only within your organization, give
+ non-standard names with instructions, or make other arrangements.\n\n 3.
+ Distributing in object code or binary form, with similar options: distribute
+ with instructions, include source, use non-standard names, or other arrangements.\n\n 4.
+ Modifying and including parts in other software, but some files have different
+ conditions (refer to LEGAL file).\n\n 5. Scripts and library files generated
+ by the software don''t fall under this copyright.\n\n 6. Standard disclaimer:
+ software is provided \"AS IS\" with no warranties.\n\nThis seems to be a dual-licensed
+ approach: either under the 2-clause BSDL or these specific terms.\n\nMy response
+ should be helpful, accurate, and not verbose. I need to explain what''s in
+ the file without copying it verbatim, as per the instruction to keep it clear
+ and concise.\n\nStructure my response:\n\n1. **Introduction:** State that
+ the file is a license for Ruby.\n\n2. **Summary of key points:** Highlight
+ the main permissions and conditions.\n\n3. **Mention alternatives:** Note
+ the option for 2-clause BSDL.\n\n4. **End with disclaimer:** Include the warranty
+ disclaimer part.\n\nSince the user might not be familiar, I should use simple
+ language.\n\nFinally, ensure my response is neutral and informative.","refusal":null},"finish_reason":"stop"}],"usage":{"prompt_tokens":531,"completion_tokens":392,"total_tokens":1366,"prompt_tokens_details":{"text_tokens":531,"audio_tokens":0,"image_tokens":0,"cached_tokens":530},"completion_tokens_details":{"reasoning_tokens":443,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 10:00:09 GMT
+recorded_with: VCR 6.3.1
diff --git a/spec/fixtures/vcr_cassettes/chat_text_models_xai_grok-3-mini_can_understand_text.yml b/spec/fixtures/vcr_cassettes/chat_text_models_xai_grok-3-mini_can_understand_text.yml
new file mode 100644
index 000000000..0708b20f0
--- /dev/null
+++ b/spec/fixtures/vcr_cassettes/chat_text_models_xai_grok-3-mini_can_understand_text.yml
@@ -0,0 +1,146 @@
+---
+http_interactions:
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":[{"type":"text","text":"What''s
+ in this file?"},{"type":"text","text":"Ruby
+ is the best."}]}],"stream":false}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:59:44 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: !binary |-
+ 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
+ recorded_at: Thu, 04 Sep 2025 09:59:44 GMT
+- request:
+ method: post
+ uri: https://api.x.ai/v1/chat/completions
+ body:
+ encoding: UTF-8
+ string: '{"model":"grok-3-mini","messages":[{"role":"user","content":[{"type":"text","text":"What''s
+ in this file?"},{"type":"text","text":"Ruby
+ is the best."}]},{"role":"assistant","content":"The file \"ruby.txt\"
+ is a plain text file (as indicated by its MIME type ''text/plain''), and its
+ content is:\n\n```\nRuby is the best.\n```"},{"role":"user","content":[{"type":"text","text":"and
+ in this one?"},{"type":"text","text":"Ruby
+ is the best"}]}],"stream":false}'
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 09:59:53 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Zero-Data-Retention:
+ - 'false'
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"id":"ee5313b2-3fe3-52a2-1ec7-adc020a77ab2_us-east-1","object":"chat.completion","created":1756979984,"model":"grok-3-mini","choices":[{"index":0,"message":{"role":"assistant","content":"The
+ file \"ruby.xml\" is an XML file (as indicated by its MIME type ''application/xml''),
+ and its content is:\n\n```\nRuby is the best\n```\n\nThis
+ appears to be a simple XML document with a single element named \"truism\"
+ containing the text \"Ruby is the best\". Note that XML files typically require
+ a proper structure, such as a root element and possibly an XML declaration
+ (e.g., ), but based on the provided information, this
+ is the content as given.","reasoning_content":"First, the user is asking about
+ another file, named ''ruby.xml'' with MIME type ''application/xml''. The content
+ provided is:\n\nRuby
+ is the best\n\nThis looks like an XML file. In XML, the structure
+ is defined by tags. So, this appears to be a simple XML document with a root
+ element or something similar.\n\nThe content inside is: Ruby is the
+ best\n\nIn the message, it''s wrapped in tags, but I think
+ the actual content of the file is what''s inside: Ruby is the best\n\nSo,
+ the file ''ruby.xml'' contains XML data: an element named \"truism\" with
+ the text \"Ruby is the best\".\n\nMy response should be similar to the previous
+ one, where I described the file''s content based on the provided information.\n\nIn
+ the previous interaction, for ''ruby.txt'', I said:\n\n\"The file \"ruby.txt\"
+ is a plain text file... and its content is: Ruby is the best.\"\n\nSo, for
+ consistency, I should say something like:\n\n\"The file \"ruby.xml\" is an
+ XML file (as indicated by its MIME type ''application/xml''), and its content
+ is: Ruby is the best\"\n\nI need to be precise. The full
+ content might include the XML declaration if it''s a proper XML file, but
+ it''s not specified here. Based on what''s given, I''ll stick to what''s provided.\n\nIn
+ the user''s message, it''s enclosed in tags, but I should extract the
+ inner content as the file''s content.\n\nFinally, structure my response clearly.","refusal":null},"finish_reason":"stop"}],"usage":{"prompt_tokens":103,"completion_tokens":114,"total_tokens":565,"prompt_tokens_details":{"text_tokens":103,"audio_tokens":0,"image_tokens":0,"cached_tokens":32},"completion_tokens_details":{"reasoning_tokens":348,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0},"num_sources_used":0},"system_fingerprint":"fp_6a09108ff5"}'
+ recorded_at: Thu, 04 Sep 2025 09:59:53 GMT
+recorded_with: VCR 6.3.1
From 7bafed505593619c627981c90cc0fe2e755aa05c Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Thu, 4 Sep 2025 11:10:46 +0100
Subject: [PATCH 23/28] Add xAI model VCRs
---
...r_ollama_without_warning_after_refresh.yml | 104 ++++++++++++++++++
...nd_returns_a_chainable_models_instance.yml | 104 ++++++++++++++++++
...ls_refresh_works_as_a_class_method_too.yml | 104 ++++++++++++++++++
3 files changed, 312 insertions(+)
diff --git a/spec/fixtures/vcr_cassettes/models_local_provider_model_fetching_local_provider_model_resolution_assumes_model_exists_for_ollama_without_warning_after_refresh.yml b/spec/fixtures/vcr_cassettes/models_local_provider_model_fetching_local_provider_model_resolution_assumes_model_exists_for_ollama_without_warning_after_refresh.yml
index 4d035fd1c..f052f8393 100644
--- a/spec/fixtures/vcr_cassettes/models_local_provider_model_fetching_local_provider_model_resolution_assumes_model_exists_for_ollama_without_warning_after_refresh.yml
+++ b/spec/fixtures/vcr_cassettes/models_local_provider_model_fetching_local_provider_model_resolution_assumes_model_exists_for_ollama_without_warning_after_refresh.yml
@@ -2142,6 +2142,110 @@ http_interactions:
]
}
recorded_at: Wed, 03 Sep 2025 15:02:18 GMT
+- request:
+ method: get
+ uri: https://api.x.ai/v1/language-models
+ body:
+ encoding: US-ASCII
+ string: ''
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 08:54:14 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"models":[{"id":"grok-2-1212","fingerprint":"fp_fced99f24f","created":1737331200,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":20000,"cached_prompt_text_token_price":0,"prompt_image_token_price":0,"completion_text_token_price":100000,"search_price":0,"aliases":["grok-2","grok-2-latest"]},{"id":"grok-2-vision-1212","fingerprint":"fp_236db45c03","created":1733961600,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text","image"],"output_modalities":["text"],"prompt_text_token_price":20000,"cached_prompt_text_token_price":0,"prompt_image_token_price":20000,"completion_text_token_price":100000,"search_price":0,"aliases":["grok-2-vision","grok-2-vision-latest"]},{"id":"grok-3","fingerprint":"fp_898ae9f31c","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":30000,"cached_prompt_text_token_price":7500,"prompt_image_token_price":0,"completion_text_token_price":150000,"search_price":250000000,"aliases":["grok-3-latest","grok-3-beta"]},{"id":"grok-3-fast","fingerprint":"fp_9c1fa52e52","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":50000,"cached_prompt_text_token_price":12500,"prompt_image_token_price":0,"completion_text_token_price":250000,"search_price":250000000,"aliases":["grok-3-fast-latest","grok-3-fast-beta"]},{"id":"grok-3-mini","fingerprint":"fp_6a09108ff5","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":3000,"cached_prompt_text_token_price":750,"prompt_image_token_price":0,"completion_text_token_price":5000,"search_price":250000000,"aliases":["grok-3-mini-latest","grok-3-mini-beta"]},{"id":"grok-3-mini-fast","fingerprint":"fp_5097008838","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":6000,"cached_prompt_text_token_price":1500,"prompt_image_token_price":0,"completion_text_token_price":40000,"search_price":250000000,"aliases":["grok-3-mini-fast-latest","grok-3-mini-fast-beta"]},{"id":"grok-4-0709","fingerprint":"fp_19e21a36c0","created":1752019200,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":30000,"cached_prompt_text_token_price":7500,"prompt_image_token_price":30000,"completion_text_token_price":150000,"search_price":250000000,"aliases":["grok-4","grok-4-latest"]},{"id":"grok-4-0709-eu","fingerprint":"fp_e83149a648","created":1752019200,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text","image"],"output_modalities":["text"],"prompt_text_token_price":30000,"cached_prompt_text_token_price":7500,"prompt_image_token_price":30000,"completion_text_token_price":150000,"search_price":250000000,"aliases":[]},{"id":"grok-code-fast-1","fingerprint":"fp_10f00c862d","created":1755993600,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":2000,"cached_prompt_text_token_price":200,"prompt_image_token_price":0,"completion_text_token_price":15000,"search_price":0,"aliases":["grok-code-fast","grok-code-fast-1-0825"]}]}'
+ recorded_at: Thu, 04 Sep 2025 08:54:14 GMT
+- request:
+ method: get
+ uri: https://api.x.ai/v1/image-generation-models
+ body:
+ encoding: US-ASCII
+ string: ''
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 08:54:15 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"models":[{"id":"grok-2-image-1212","fingerprint":"fp_cef22a67b0","max_prompt_length":1024,"created":1736726400,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text","image"],"output_modalities":["image"],"image_price":700000000,"aliases":["grok-2-image","grok-2-image-latest"]}]}'
+ recorded_at: Thu, 04 Sep 2025 08:54:15 GMT
- request:
method: get
uri: https://api.parsera.org/v1/llm-specs
diff --git a/spec/fixtures/vcr_cassettes/models_refresh_updates_models_and_returns_a_chainable_models_instance.yml b/spec/fixtures/vcr_cassettes/models_refresh_updates_models_and_returns_a_chainable_models_instance.yml
index 38e342b18..ceae7ee1c 100644
--- a/spec/fixtures/vcr_cassettes/models_refresh_updates_models_and_returns_a_chainable_models_instance.yml
+++ b/spec/fixtures/vcr_cassettes/models_refresh_updates_models_and_returns_a_chainable_models_instance.yml
@@ -2148,6 +2148,110 @@ http_interactions:
]
}
recorded_at: Wed, 27 Aug 2025 16:51:20 GMT
+- request:
+ method: get
+ uri: https://api.x.ai/v1/language-models
+ body:
+ encoding: US-ASCII
+ string: ''
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 08:54:14 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"models":[{"id":"grok-2-1212","fingerprint":"fp_fced99f24f","created":1737331200,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":20000,"cached_prompt_text_token_price":0,"prompt_image_token_price":0,"completion_text_token_price":100000,"search_price":0,"aliases":["grok-2","grok-2-latest"]},{"id":"grok-2-vision-1212","fingerprint":"fp_236db45c03","created":1733961600,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text","image"],"output_modalities":["text"],"prompt_text_token_price":20000,"cached_prompt_text_token_price":0,"prompt_image_token_price":20000,"completion_text_token_price":100000,"search_price":0,"aliases":["grok-2-vision","grok-2-vision-latest"]},{"id":"grok-3","fingerprint":"fp_898ae9f31c","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":30000,"cached_prompt_text_token_price":7500,"prompt_image_token_price":0,"completion_text_token_price":150000,"search_price":250000000,"aliases":["grok-3-latest","grok-3-beta"]},{"id":"grok-3-fast","fingerprint":"fp_9c1fa52e52","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":50000,"cached_prompt_text_token_price":12500,"prompt_image_token_price":0,"completion_text_token_price":250000,"search_price":250000000,"aliases":["grok-3-fast-latest","grok-3-fast-beta"]},{"id":"grok-3-mini","fingerprint":"fp_6a09108ff5","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":3000,"cached_prompt_text_token_price":750,"prompt_image_token_price":0,"completion_text_token_price":5000,"search_price":250000000,"aliases":["grok-3-mini-latest","grok-3-mini-beta"]},{"id":"grok-3-mini-fast","fingerprint":"fp_5097008838","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":6000,"cached_prompt_text_token_price":1500,"prompt_image_token_price":0,"completion_text_token_price":40000,"search_price":250000000,"aliases":["grok-3-mini-fast-latest","grok-3-mini-fast-beta"]},{"id":"grok-4-0709","fingerprint":"fp_19e21a36c0","created":1752019200,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":30000,"cached_prompt_text_token_price":7500,"prompt_image_token_price":30000,"completion_text_token_price":150000,"search_price":250000000,"aliases":["grok-4","grok-4-latest"]},{"id":"grok-4-0709-eu","fingerprint":"fp_e83149a648","created":1752019200,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text","image"],"output_modalities":["text"],"prompt_text_token_price":30000,"cached_prompt_text_token_price":7500,"prompt_image_token_price":30000,"completion_text_token_price":150000,"search_price":250000000,"aliases":[]},{"id":"grok-code-fast-1","fingerprint":"fp_10f00c862d","created":1755993600,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":2000,"cached_prompt_text_token_price":200,"prompt_image_token_price":0,"completion_text_token_price":15000,"search_price":0,"aliases":["grok-code-fast","grok-code-fast-1-0825"]}]}'
+ recorded_at: Thu, 04 Sep 2025 08:54:14 GMT
+- request:
+ method: get
+ uri: https://api.x.ai/v1/image-generation-models
+ body:
+ encoding: US-ASCII
+ string: ''
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 08:54:15 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"models":[{"id":"grok-2-image-1212","fingerprint":"fp_cef22a67b0","max_prompt_length":1024,"created":1736726400,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text","image"],"output_modalities":["image"],"image_price":700000000,"aliases":["grok-2-image","grok-2-image-latest"]}]}'
+ recorded_at: Thu, 04 Sep 2025 08:54:15 GMT
- request:
method: get
uri: https://api.parsera.org/v1/llm-specs
diff --git a/spec/fixtures/vcr_cassettes/models_refresh_works_as_a_class_method_too.yml b/spec/fixtures/vcr_cassettes/models_refresh_works_as_a_class_method_too.yml
index d3d11be81..f203800cd 100644
--- a/spec/fixtures/vcr_cassettes/models_refresh_works_as_a_class_method_too.yml
+++ b/spec/fixtures/vcr_cassettes/models_refresh_works_as_a_class_method_too.yml
@@ -2148,6 +2148,110 @@ http_interactions:
]
}
recorded_at: Wed, 27 Aug 2025 16:51:24 GMT
+- request:
+ method: get
+ uri: https://api.x.ai/v1/language-models
+ body:
+ encoding: US-ASCII
+ string: ''
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 08:54:14 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"models":[{"id":"grok-2-1212","fingerprint":"fp_fced99f24f","created":1737331200,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":20000,"cached_prompt_text_token_price":0,"prompt_image_token_price":0,"completion_text_token_price":100000,"search_price":0,"aliases":["grok-2","grok-2-latest"]},{"id":"grok-2-vision-1212","fingerprint":"fp_236db45c03","created":1733961600,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text","image"],"output_modalities":["text"],"prompt_text_token_price":20000,"cached_prompt_text_token_price":0,"prompt_image_token_price":20000,"completion_text_token_price":100000,"search_price":0,"aliases":["grok-2-vision","grok-2-vision-latest"]},{"id":"grok-3","fingerprint":"fp_898ae9f31c","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":30000,"cached_prompt_text_token_price":7500,"prompt_image_token_price":0,"completion_text_token_price":150000,"search_price":250000000,"aliases":["grok-3-latest","grok-3-beta"]},{"id":"grok-3-fast","fingerprint":"fp_9c1fa52e52","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":50000,"cached_prompt_text_token_price":12500,"prompt_image_token_price":0,"completion_text_token_price":250000,"search_price":250000000,"aliases":["grok-3-fast-latest","grok-3-fast-beta"]},{"id":"grok-3-mini","fingerprint":"fp_6a09108ff5","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":3000,"cached_prompt_text_token_price":750,"prompt_image_token_price":0,"completion_text_token_price":5000,"search_price":250000000,"aliases":["grok-3-mini-latest","grok-3-mini-beta"]},{"id":"grok-3-mini-fast","fingerprint":"fp_5097008838","created":1743724800,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":6000,"cached_prompt_text_token_price":1500,"prompt_image_token_price":0,"completion_text_token_price":40000,"search_price":250000000,"aliases":["grok-3-mini-fast-latest","grok-3-mini-fast-beta"]},{"id":"grok-4-0709","fingerprint":"fp_19e21a36c0","created":1752019200,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":30000,"cached_prompt_text_token_price":7500,"prompt_image_token_price":30000,"completion_text_token_price":150000,"search_price":250000000,"aliases":["grok-4","grok-4-latest"]},{"id":"grok-4-0709-eu","fingerprint":"fp_e83149a648","created":1752019200,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text","image"],"output_modalities":["text"],"prompt_text_token_price":30000,"cached_prompt_text_token_price":7500,"prompt_image_token_price":30000,"completion_text_token_price":150000,"search_price":250000000,"aliases":[]},{"id":"grok-code-fast-1","fingerprint":"fp_10f00c862d","created":1755993600,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text"],"output_modalities":["text"],"prompt_text_token_price":2000,"cached_prompt_text_token_price":200,"prompt_image_token_price":0,"completion_text_token_price":15000,"search_price":0,"aliases":["grok-code-fast","grok-code-fast-1-0825"]}]}'
+ recorded_at: Thu, 04 Sep 2025 08:54:14 GMT
+- request:
+ method: get
+ uri: https://api.x.ai/v1/image-generation-models
+ body:
+ encoding: US-ASCII
+ string: ''
+ headers:
+ User-Agent:
+ - Faraday v2.13.1
+ Authorization:
+ - Bearer
+ Content-Type:
+ - application/json
+ Accept-Encoding:
+ - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+ Accept:
+ - "*/*"
+ response:
+ status:
+ code: 200
+ message: OK
+ headers:
+ Date:
+ - Thu, 04 Sep 2025 08:54:15 GMT
+ Content-Type:
+ - application/json
+ Transfer-Encoding:
+ - chunked
+ Connection:
+ - keep-alive
+ X-Request-Id:
+ - ""
+ Vary:
+ - origin, access-control-request-method, access-control-request-headers
+ Access-Control-Allow-Origin:
+ - "*"
+ Access-Control-Expose-Headers:
+ - "*"
+ Cf-Cache-Status:
+ - DYNAMIC
+ Set-Cookie:
+ - ""
+ Strict-Transport-Security:
+ - max-age=31536000
+ Server:
+ - cloudflare
+ Cf-Ray:
+ - ""
+ body:
+ encoding: ASCII-8BIT
+ string: '{"models":[{"id":"grok-2-image-1212","fingerprint":"fp_cef22a67b0","max_prompt_length":1024,"created":1736726400,"object":"model","owned_by":"xai","version":"1.0","input_modalities":["text","image"],"output_modalities":["image"],"image_price":700000000,"aliases":["grok-2-image","grok-2-image-latest"]}]}'
+ recorded_at: Thu, 04 Sep 2025 08:54:15 GMT
- request:
method: get
uri: https://api.parsera.org/v1/llm-specs
From 45f61c1b28e2ecfbdad8adaa8c91219d77f11641 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Thu, 4 Sep 2025 11:18:15 +0100
Subject: [PATCH 24/28] Add xAI API key to docs
---
docs/_getting_started/configuration.md | 1 +
1 file changed, 1 insertion(+)
diff --git a/docs/_getting_started/configuration.md b/docs/_getting_started/configuration.md
index 6af299ca4..0d8e86306 100644
--- a/docs/_getting_started/configuration.md
+++ b/docs/_getting_started/configuration.md
@@ -59,6 +59,7 @@ RubyLLM.configure do |config|
config.mistral_api_key = ENV['MISTRAL_API_KEY']
config.perplexity_api_key = ENV['PERPLEXITY_API_KEY']
config.openrouter_api_key = ENV['OPENROUTER_API_KEY']
+ config.xai_api_key = ENV['XAI_API_KEY']
# Local providers
config.ollama_api_base = 'http://localhost:11434/v1'
From 163f9d0118e37a5d04e27281911621d3aef79e80 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Thu, 4 Sep 2025 11:33:08 +0100
Subject: [PATCH 25/28] Refactor Provider.list_models
---
lib/ruby_llm/provider.rb | 11 +++--------
lib/ruby_llm/providers/xai/models.rb | 4 +---
2 files changed, 4 insertions(+), 11 deletions(-)
diff --git a/lib/ruby_llm/provider.rb b/lib/ruby_llm/provider.rb
index 2e6d41409..51eff48ee 100644
--- a/lib/ruby_llm/provider.rb
+++ b/lib/ruby_llm/provider.rb
@@ -60,14 +60,9 @@ def complete(messages, tools:, temperature:, model:, params: {}, headers: {}, sc
end
def list_models
- if models_url.is_a?(Array)
- models_url.map do |url|
- response = @connection.get(url)
- parse_list_models_response response, slug, capabilities
- end.flatten
- else
- response = @connection.get models_url
- parse_list_models_response response, slug, capabilities
+ Array(models_url).flat_map do |url|
+ response = @connection.get(url)
+ parse_list_models_response(response, slug, capabilities)
end
end
diff --git a/lib/ruby_llm/providers/xai/models.rb b/lib/ruby_llm/providers/xai/models.rb
index 574c120bc..ab52d83a1 100644
--- a/lib/ruby_llm/providers/xai/models.rb
+++ b/lib/ruby_llm/providers/xai/models.rb
@@ -24,7 +24,7 @@ def parse_list_models_response(response, slug, capabilities)
data['models']&.map do |model_data|
model_id = model_data['id']
- info = Model::Info.new(
+ Model::Info.new(
id: model_id,
name: capabilities.format_display_name(model_id),
provider: slug,
@@ -40,8 +40,6 @@ def parse_list_models_response(response, slug, capabilities)
aliases: model_data['aliases']
}
)
- puts info.inspect
- info
end || []
end
end
From 293d109cfed73e2b1645568e5633dd258d0c4bfd Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Sun, 7 Sep 2025 11:33:15 +0100
Subject: [PATCH 26/28] Improve xAI skip message in can use tools without
parameters in multi-turn streaming conversations spec
---
spec/ruby_llm/chat_tools_spec.rb | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/spec/ruby_llm/chat_tools_spec.rb b/spec/ruby_llm/chat_tools_spec.rb
index c9390c299..0e16db0ba 100644
--- a/spec/ruby_llm/chat_tools_spec.rb
+++ b/spec/ruby_llm/chat_tools_spec.rb
@@ -142,7 +142,10 @@ def execute(query:)
skip 'Mistral has a bug with tool arguments in multi-turn streaming' if provider == :mistral
- skip 'xAI has a bug with tool arguments in multi-turn streaming' if provider == :xai
+ if provider == :xai
+ skip 'xAI model infinitely loops calling tools when tool has no parameters' \
+ 'and always provides the same result in multi-turn streaming conversations'
+ end
unless RubyLLM::Provider.providers[provider]&.local?
model_info = RubyLLM.models.find(model)
From b1b7cbfb18083de240debbd779a671b3f1c2c62e Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Fri, 26 Sep 2025 05:39:26 +0200
Subject: [PATCH 27/28] Revert list_models change and override in xAI Provider
---
lib/ruby_llm/provider.rb | 6 ++----
lib/ruby_llm/providers/xai.rb | 9 +++++++++
2 files changed, 11 insertions(+), 4 deletions(-)
diff --git a/lib/ruby_llm/provider.rb b/lib/ruby_llm/provider.rb
index 51eff48ee..2d108cd61 100644
--- a/lib/ruby_llm/provider.rb
+++ b/lib/ruby_llm/provider.rb
@@ -60,10 +60,8 @@ def complete(messages, tools:, temperature:, model:, params: {}, headers: {}, sc
end
def list_models
- Array(models_url).flat_map do |url|
- response = @connection.get(url)
- parse_list_models_response(response, slug, capabilities)
- end
+ response = @connection.get models_url
+ parse_list_models_response response, slug, capabilities
end
def embed(text, model:, dimensions:)
diff --git a/lib/ruby_llm/providers/xai.rb b/lib/ruby_llm/providers/xai.rb
index b77015c7a..bc50e9cd7 100644
--- a/lib/ruby_llm/providers/xai.rb
+++ b/lib/ruby_llm/providers/xai.rb
@@ -19,6 +19,15 @@ def headers
}
end
+ # xAI uses separate endpoints for langauge and image models.
+ # Override Provider class method here to support multiple model URLs.
+ def list_models
+ Array(models_url).flat_map do |url|
+ response = @connection.get(url)
+ parse_list_models_response(response, slug, capabilities)
+ end
+ end
+
# xAI uses a different error format than OpenAI
# {"code": "...", "error": "..."}
def parse_error(response)
From e0d53b8523f557c45b66473f77fdb91e187ea162 Mon Sep 17 00:00:00 2001
From: Ritchie Cotton <243725+infinityrobot@users.noreply.github.com>
Date: Fri, 26 Sep 2025 05:47:41 +0200
Subject: [PATCH 28/28] Update models, aliases and docs
---
docs/_reference/available-models.md | 1097 +++---
lib/ruby_llm/aliases.json | 105 +-
lib/ruby_llm/models.json | 5312 +++++++++++++++------------
3 files changed, 3621 insertions(+), 2893 deletions(-)
diff --git a/docs/_reference/available-models.md b/docs/_reference/available-models.md
index b76946b46..429d8e2f0 100644
--- a/docs/_reference/available-models.md
+++ b/docs/_reference/available-models.md
@@ -30,7 +30,7 @@ redirect_from:
## Last Updated
{: .d-inline-block }
-2025-09-21
+2025-09-26
{: .label .label-green }
## Models by Provider
@@ -39,44 +39,44 @@ redirect_from:
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
-| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
+| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00 |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| claude-opus-4-20250514 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| claude-opus-4-1-20250805 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
+| claude-opus-4-20250514 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| claude-opus-4-1-20250805 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
-| claude-3-7-sonnet-20250219 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| claude-sonnet-4-20250514 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
+| claude-3-7-sonnet-20250219 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
+| claude-sonnet-4-20250514 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
### Bedrock (28)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
-| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
-| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| us.anthropic.claude-opus-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
@@ -98,24 +98,24 @@ redirect_from:
| :-- | :-- | --: | --: | :-- |
| embedding-001 | gemini | 2048 | 1 | - |
| embedding-gecko-001 | gemini | 1024 | 1 | - |
-| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
-| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-latest | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
-| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-live-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-preview-image-generation | gemini | 32000 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
-| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-preview | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-lite-preview-02-05 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
@@ -124,35 +124,35 @@ redirect_from:
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-image-preview | gemini | 32768 | 32768 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-live-2.5-flash-preview | gemini | 1048576 | 8192 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-native-audio-preview-09-2025 | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-exp-native-audio-thinking-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.5-flash-preview-native-audio-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.5-flash-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash-preview-tts | gemini | 8000 | 16000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-preview-06-17 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-2.5-flash-lite-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro | gemini | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| gemini-2.5-pro-preview-06-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-03-25 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-05-06 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-tts | gemini | 8000 | 16000 | In: $1.25, Out: $10.00, Cache: $0.31 |
-| gemini-embedding-001 | gemini | 2048 | 1 | - |
-| gemini-embedding-exp | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-exp-1206 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-lite-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-pro-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-robotics-er-1.5-preview | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemma-3-12b-it | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemma-3-1b-it | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemma-3-27b-it | gemini | 131072 | 8192 | In: $0.08, Out: $0.30 |
| gemma-3-4b-it | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemma-3n-e2b-it | gemini | 8192 | 2048 | In: $0.08, Out: $0.30 |
| gemma-3n-e4b-it | gemini | 8192 | 2048 | In: $0.08, Out: $0.30 |
-| imagen-3.0-generate-002 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-preview-06-06 | gemini | 480 | 8192 | - |
-| imagen-4.0-ultra-generate-preview-06-06 | gemini | 480 | 8192 | - |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
-| aqa | gemini | 7168 | 1024 | - |
| text-embedding-004 | gemini | 2048 | 1 | - |
@@ -230,103 +230,114 @@ redirect_from:
| voxtral-small-latest | mistral | 32768 | 8192 | - |
-### OpenAI (90)
+### OpenAI (101)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| babbage-002 | openai | 4096 | 16384 | In: $0.40, Out: $0.40 |
| chatgpt-4o-latest | openai | 128000 | 16384 | In: $5.00, Out: $15.00 |
-| codex-mini-latest | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| dall-e-2 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| dall-e-3 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| davinci-002 | openai | 4096 | 16384 | In: $2.00, Out: $2.00 |
+| dall-e-2 | openai | - | - | - |
+| dall-e-3 | openai | - | - | - |
+| gpt-image-1 | openai | - | - | In: $5.00, Out: $40.00, Cache: $1.25 |
| gpt-3.5-turbo | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-0125 | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-1106 | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-16k | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-instruct | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-instruct-0914 | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
-| gpt-4 | openai | 8192 | 8192 | In: $10.00, Out: $30.00 |
+| gpt-4 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
+| gpt-4-0613 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
| gpt-4-0125-preview | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-4-0613 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
| gpt-4-1106-preview | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4-turbo-preview | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
-| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-audio-preview-2024-10-01 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-audio-preview-2025-06-03 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-transcribe | openai | 16000 | 2000 | In: $2.50, Out: $10.00 |
+| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-realtime-preview | openai | 128000 | 4096 | In: $0.60, Out: $2.40 |
-| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $0.60, Out: $2.40 |
-| gpt-4o-mini-search-preview | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-4o-mini-search-preview-2025-03-11 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
+| gpt-4o-mini-realtime-preview | openai | 16000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 16000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-search-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-tts | openai | 2000 | - | In: $0.60, Out: $12.00 |
| gpt-4o-mini-transcribe | openai | 16000 | 2000 | In: $1.25, Out: $5.00 |
-| gpt-4o-mini-tts | openai | - | - | In: $0.60, Out: $12.00 |
+| gpt-4o-audio-preview-2024-10-01 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-audio-preview-2025-06-03 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-realtime-preview | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2024-10-01 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2025-06-03 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
-| gpt-4o-transcribe | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-5 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-chat-latest | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-audio | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-audio-2025-08-28 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-image-1 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-realtime | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-realtime-2025-08-28 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-mini | openai | 128000 | 65536 | In: $1.10, Out: $4.40 |
-| o1-mini-2024-09-12 | openai | 128000 | 65536 | In: $1.10, Out: $4.40 |
-| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
-| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
-| o3 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-2025-04-16 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-deep-research | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-deep-research-2025-06-26 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40 |
-| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40 |
-| o3-pro | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-pro-2025-06-10 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o4-mini | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o4-mini-2025-04-16 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o4-mini-deep-research | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o4-mini-deep-research-2025-06-26 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
+| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-chat-latest | openai | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-codex | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| omni-moderation-2024-09-26 | openai | - | - | - |
-| omni-moderation-latest | openai | - | - | - |
-| tts-1 | openai | - | - | In: $15.00, Out: $15.00 |
+| tts-1 | openai | - | - | In: $15.00 |
| tts-1-1106 | openai | - | - | In: $15.00, Out: $15.00 |
-| tts-1-hd | openai | - | - | In: $30.00, Out: $30.00 |
+| tts-1-hd | openai | - | - | In: $30.00 |
| tts-1-hd-1106 | openai | - | - | In: $30.00, Out: $30.00 |
-| whisper-1 | openai | - | - | In: $0.01, Out: $0.01 |
-| text-embedding-3-large | openai | - | - | In: $0.13, Out: $0.13 |
-| text-embedding-3-small | openai | - | - | In: $0.02, Out: $0.02 |
-| text-embedding-ada-002 | openai | - | - | In: $0.10, Out: $0.10 |
+| whisper-1 | openai | - | - | In: $0.01 |
+| babbage-002 | openai | - | 16384 | In: $0.40, Out: $0.40 |
+| codex-mini-latest | openai | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
+| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| computer-use-preview-2025-03-11 | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| davinci-002 | openai | - | 16384 | In: $2.00, Out: $2.00 |
+| gpt-audio | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-audio-2025-08-28 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-oss-120b | openai | 131072 | 131072 | - |
+| gpt-oss-20b | openai | 131072 | 131072 | - |
+| gpt-realtime-2025-08-28 | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
+| gpt-realtime | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
+| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-preview | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-preview-2024-09-12 | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-mini-2024-09-12 | openai | 128000 | 65536 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o1-mini | openai | 128000 | 65536 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-deep-research | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-pro | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o3-pro-2025-06-10 | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o4-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini-2025-04-16 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o4-mini-deep-research | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| omni-moderation-latest | openai | - | - | - |
+| text-embedding-3-large | openai | - | - | In: $0.13 |
+| text-embedding-3-small | openai | - | - | In: $0.02 |
+| text-embedding-ada-002 | openai | - | - | In: $0.10 |
+| text-moderation-latest | openai | - | 32768 | - |
+| text-moderation-stable | openai | - | 32768 | - |
-### OpenRouter (327)
+### OpenRouter (325)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
@@ -369,13 +380,8 @@ redirect_from:
| bytedance/seed-oss-36b-instruct | openrouter | 131072 | - | In: $0.16, Out: $0.65 |
| bytedance/ui-tars-1.5-7b | openrouter | 128000 | 2048 | In: $0.10, Out: $0.20 |
| deepcogito/cogito-v2-preview-llama-109b-moe | openrouter | 32767 | - | In: $0.18, Out: $0.59 |
-| cohere/command | openrouter | 4096 | 4000 | In: $1.00, Out: $2.00 |
| cohere/command-a | openrouter | 256000 | 8192 | In: $2.50, Out: $10.00 |
-| cohere/command-r | openrouter | 128000 | 4000 | In: $0.50, Out: $1.50 |
-| cohere/command-r-03-2024 | openrouter | 128000 | 4000 | In: $0.50, Out: $1.50 |
| cohere/command-r-08-2024 | openrouter | 128000 | 4000 | In: $0.15, Out: $0.60 |
-| cohere/command-r-plus | openrouter | 128000 | 4000 | In: $3.00, Out: $15.00 |
-| cohere/command-r-plus-04-2024 | openrouter | 128000 | 4000 | In: $3.00, Out: $15.00 |
| cohere/command-r-plus-08-2024 | openrouter | 128000 | 4000 | In: $2.50, Out: $10.00 |
| cohere/command-r7b-12-2024 | openrouter | 128000 | 4000 | In: $0.04, Out: $0.15 |
| deepcogito/cogito-v2-preview-deepseek-671b | openrouter | 163840 | - | In: $1.25, Out: $1.25 |
@@ -386,6 +392,7 @@ redirect_from:
| deepseek/deepseek-chat-v3.1 | openrouter | 163840 | - | In: $0.25, Out: $1.00 |
| deepseek/deepseek-chat-v3.1:free | openrouter | 163840 | - | - |
| deepseek/deepseek-v3.1-base | openrouter | 163840 | - | In: $0.25, Out: $1.00 |
+| deepseek/deepseek-v3.1-terminus | openrouter | 163840 | - | In: $0.27, Out: $1.00 |
| deepseek/deepseek-r1-0528-qwen3-8b | openrouter | 131072 | - | In: $0.01, Out: $0.05 |
| deepseek/deepseek-r1-0528-qwen3-8b:free | openrouter | 131072 | - | - |
| deepseek/deepseek-r1 | openrouter | 163840 | 163840 | In: $0.40, Out: $2.00 |
@@ -402,17 +409,17 @@ redirect_from:
| cognitivecomputations/dolphin3.0-r1-mistral-24b | openrouter | 32768 | - | In: $0.01, Out: $0.03 |
| cognitivecomputations/dolphin3.0-r1-mistral-24b:free | openrouter | 32768 | - | - |
| eleutherai/llemma_7b | openrouter | 4096 | 4096 | In: $0.80, Out: $1.20 |
+| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
| alpindale/goliath-120b | openrouter | 6144 | 512 | In: $4.00, Out: $5.50 |
-| google/gemini-flash-1.5 | openrouter | 1000000 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| google/gemini-flash-1.5-8b | openrouter | 1000000 | 8192 | In: $0.04, Out: $0.15, Cache: $0.01 |
-| google/gemini-pro-1.5 | openrouter | 2000000 | 8192 | In: $1.25, Out: $5.00 |
| google/gemini-2.0-flash-001 | openrouter | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.0-flash-exp:free | openrouter | 1048576 | 8192 | - |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| google/gemini-2.5-flash-lite-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| google/gemini-2.5-flash-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview-05-06 | openrouter | 1048576 | 65535 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -436,14 +443,13 @@ redirect_from:
| liquid/lfm-7b | openrouter | 32768 | - | In: $0.01, Out: $0.01 |
| meta-llama/llama-guard-3-8b | openrouter | 131072 | - | In: $0.02, Out: $0.06 |
| anthracite-org/magnum-v2-72b | openrouter | 32768 | - | In: $3.00, Out: $3.00 |
-| anthracite-org/magnum-v4-72b | openrouter | 16384 | 2048 | In: $2.00, Out: $5.00 |
+| anthracite-org/magnum-v4-72b | openrouter | 16384 | 2048 | In: $2.50, Out: $5.00 |
| mancer/weaver | openrouter | 8000 | 2000 | In: $1.12, Out: $1.12 |
| meituan/longcat-flash-chat | openrouter | 131072 | - | In: $0.12, Out: $0.60 |
| meta-llama/llama-3-70b-instruct | openrouter | 8192 | 16384 | In: $0.30, Out: $0.40 |
| meta-llama/llama-3-8b-instruct | openrouter | 8192 | 16384 | In: $0.03, Out: $0.06 |
| meta-llama/llama-3.1-405b | openrouter | 32768 | - | In: $2.00, Out: $2.00 |
| meta-llama/llama-3.1-405b-instruct | openrouter | 32768 | 16384 | In: $0.80, Out: $0.80 |
-| meta-llama/llama-3.1-405b-instruct:free | openrouter | 65536 | - | - |
| meta-llama/llama-3.1-70b-instruct | openrouter | 131072 | 16384 | In: $0.10, Out: $0.28 |
| meta-llama/llama-3.1-8b-instruct | openrouter | 16384 | 16384 | In: $0.02, Out: $0.03 |
| meta-llama/llama-3.2-11b-vision-instruct | openrouter | 131072 | 16384 | In: $0.05, Out: $0.05 |
@@ -451,7 +457,7 @@ redirect_from:
| meta-llama/llama-3.2-3b-instruct | openrouter | 16384 | 16384 | In: $0.02, Out: $0.02 |
| meta-llama/llama-3.2-3b-instruct:free | openrouter | 131072 | - | - |
| meta-llama/llama-3.2-90b-vision-instruct | openrouter | 32768 | 16384 | In: $0.35, Out: $0.40 |
-| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 131072 | In: $0.01, Out: $0.04 |
+| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 131072 | In: $0.04, Out: $0.12 |
| meta-llama/llama-3.3-70b-instruct:free | openrouter | 65536 | - | - |
| meta-llama/llama-3.3-8b-instruct:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-maverick | openrouter | 1048576 | 16384 | In: $0.15, Out: $0.60 |
@@ -522,6 +528,7 @@ redirect_from:
| neversleep/llama-3-lumimaid-70b | openrouter | 8192 | 4096 | In: $4.00, Out: $6.00 |
| neversleep/llama-3.1-lumimaid-8b | openrouter | 32768 | - | In: $0.09, Out: $0.60 |
| neversleep/noromaid-20b | openrouter | 4096 | - | In: $1.00, Out: $1.75 |
+| nousresearch/deephermes-3-llama-3-8b-preview | openrouter | 131072 | - | In: $0.01, Out: $0.05 |
| nousresearch/deephermes-3-llama-3-8b-preview:free | openrouter | 131072 | - | - |
| nousresearch/deephermes-3-mistral-24b-preview | openrouter | 32768 | - | In: $0.13, Out: $0.51 |
| nousresearch/hermes-3-llama-3.1-405b | openrouter | 131072 | 131072 | In: $0.80, Out: $0.80 |
@@ -553,8 +560,9 @@ redirect_from:
| openai/gpt-4o-mini | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| openai/gpt-4o-mini-2024-07-18 | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| openai/gpt-4o-mini-search-preview | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| openai/gpt-5 | openrouter | 400000 | 128000 | In: $0.62, Out: $5.00, Cache: $0.06 |
+| openai/gpt-5 | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| openai/gpt-5-chat | openrouter | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| openai/gpt-5-codex | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| openai/gpt-5-mini | openrouter | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| openai/gpt-5-nano | openrouter | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| openai/gpt-oss-120b | openrouter | 131072 | - | In: $0.05, Out: $0.25 |
@@ -584,8 +592,6 @@ redirect_from:
| qwen/qwen-2.5-coder-32b-instruct | openrouter | 32768 | 16384 | In: $0.06, Out: $0.15 |
| qwen/qwen-2.5-coder-32b-instruct:free | openrouter | 32768 | - | - |
| qwen/qwq-32b | openrouter | 32768 | - | In: $0.15, Out: $0.40 |
-| qwen/qwq-32b:free | openrouter | 32768 | - | - |
-| qwen/qwq-32b-preview | openrouter | 32768 | - | In: $0.20, Out: $0.20 |
| qwen/qwen-plus-2025-07-28 | openrouter | 1000000 | 32768 | In: $0.40, Out: $1.20 |
| qwen/qwen-plus-2025-07-28:thinking | openrouter | 1000000 | 32768 | In: $0.40, Out: $4.00 |
| qwen/qwen-vl-max | openrouter | 7500 | 1500 | In: $0.80, Out: $3.20 |
@@ -596,13 +602,13 @@ redirect_from:
| qwen/qwen2.5-vl-32b-instruct | openrouter | 16384 | - | In: $0.04, Out: $0.14 |
| qwen/qwen2.5-vl-32b-instruct:free | openrouter | 8192 | - | - |
| qwen/qwen2.5-vl-72b-instruct | openrouter | 32768 | - | In: $0.07, Out: $0.28 |
-| qwen/qwen2.5-vl-72b-instruct:free | openrouter | 32768 | - | - |
+| qwen/qwen2.5-vl-72b-instruct:free | openrouter | 131072 | 2048 | - |
| qwen/qwen-2.5-vl-7b-instruct | openrouter | 32768 | - | In: $0.20, Out: $0.20 |
-| qwen/qwen3-14b | openrouter | 40960 | 40960 | In: $0.06, Out: $0.24 |
+| qwen/qwen3-14b | openrouter | 40960 | - | In: $0.04, Out: $0.14 |
| qwen/qwen3-14b:free | openrouter | 40960 | - | - |
| qwen/qwen3-235b-a22b | openrouter | 40960 | 40960 | In: $0.18, Out: $0.54 |
| qwen/qwen3-235b-a22b:free | openrouter | 131072 | - | - |
-| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | 262144 | In: $0.10, Out: $0.10 |
+| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | - | In: $0.09, Out: $0.60 |
| qwen/qwen3-235b-a22b-thinking-2507 | openrouter | 262144 | - | In: $0.10, Out: $0.39 |
| qwen/qwen3-30b-a3b | openrouter | 40960 | - | In: $0.06, Out: $0.22 |
| qwen/qwen3-30b-a3b:free | openrouter | 40960 | - | - |
@@ -620,6 +626,8 @@ redirect_from:
| qwen/qwen3-max | openrouter | 256000 | 32768 | In: $1.20, Out: $6.00, Cache: $0.24 |
| qwen/qwen3-next-80b-a3b-instruct | openrouter | 262144 | - | In: $0.10, Out: $0.80 |
| qwen/qwen3-next-80b-a3b-thinking | openrouter | 262144 | - | In: $0.10, Out: $0.80 |
+| qwen/qwen3-vl-235b-a22b-instruct | openrouter | 131072 | 32768 | In: $0.30, Out: $1.50 |
+| qwen/qwen3-vl-235b-a22b-thinking | openrouter | 131072 | 32768 | In: $0.30, Out: $3.00 |
| undi95/remm-slerp-l2-13b | openrouter | 6144 | - | In: $0.45, Out: $0.65 |
| sao10k/l3-lunaris-8b | openrouter | 8192 | - | In: $0.04, Out: $0.05 |
| sao10k/l3.1-euryale-70b | openrouter | 32768 | - | In: $0.65, Out: $0.75 |
@@ -637,7 +645,7 @@ redirect_from:
| tngtech/deepseek-r1t2-chimera:free | openrouter | 163840 | - | - |
| tencent/hunyuan-a13b-instruct | openrouter | 32768 | - | In: $0.03, Out: $0.03 |
| tencent/hunyuan-a13b-instruct:free | openrouter | 32768 | - | - |
-| thedrummer/anubis-70b-v1.1 | openrouter | 16384 | - | In: $0.40, Out: $0.70 |
+| thedrummer/anubis-70b-v1.1 | openrouter | 4096 | - | In: $0.40, Out: $0.70 |
| thedrummer/anubis-pro-105b-v1 | openrouter | 131072 | 131072 | In: $0.50, Out: $1.00 |
| thedrummer/rocinante-12b | openrouter | 32768 | - | In: $0.17, Out: $0.43 |
| thedrummer/skyfall-36b-v2 | openrouter | 32768 | - | In: $0.04, Out: $0.16 |
@@ -646,7 +654,7 @@ redirect_from:
| cognitivecomputations/dolphin-mistral-24b-venice-edition:free | openrouter | 32768 | - | - |
| microsoft/wizardlm-2-8x22b | openrouter | 65536 | 65536 | In: $0.48, Out: $0.48 |
| z-ai/glm-4-32b | openrouter | 128000 | - | In: $0.10, Out: $0.10 |
-| z-ai/glm-4.5 | openrouter | 131072 | - | In: $0.41, Out: $1.65 |
+| z-ai/glm-4.5 | openrouter | 131072 | - | In: $0.40, Out: $1.60 |
| z-ai/glm-4.5-air | openrouter | 131072 | 131072 | In: $0.14, Out: $0.86 |
| z-ai/glm-4.5-air:free | openrouter | 131072 | - | - |
| z-ai/glm-4.5v | openrouter | 65536 | 65536 | In: $0.50, Out: $1.80 |
@@ -655,6 +663,7 @@ redirect_from:
| x-ai/grok-3-mini | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-3-mini-beta | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| x-ai/grok-4-fast | openrouter | 2000000 | 30000 | In: $0.20, Out: $0.50, Cache: $0.05 |
| x-ai/grok-4-fast:free | openrouter | 2000000 | 30000 | - |
| x-ai/grok-code-fast-1 | openrouter | 256000 | 10000 | In: $0.20, Out: $1.50, Cache: $0.02 |
@@ -670,7 +679,7 @@ redirect_from:
| sonar-reasoning-pro | perplexity | 128000 | 8192 | In: $2.00, Out: $8.00 |
-### VertexAI (34)
+### VertexAI (20)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
@@ -686,52 +695,50 @@ redirect_from:
| gemini-2.5-flash | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-pro | vertexai | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
-| chat-bison | vertexai | - | - | - |
-| code-bison | vertexai | - | - | - |
-| code-gecko | vertexai | - | - | - |
-| codechat-bison | vertexai | - | - | - |
| gemini-embedding-001 | vertexai | - | - | - |
| gemini-exp-1121 | vertexai | - | - | - |
| gemini-exp-1206 | vertexai | - | - | - |
| gemini-pro | vertexai | - | - | - |
| gemini-pro-vision | vertexai | - | - | - |
-| image-segmentation-001 | vertexai | - | - | - |
-| imagegeneration | vertexai | - | - | - |
-| imagen-4.0-fast-generate-001 | vertexai | - | - | - |
-| imagen-4.0-generate-001 | vertexai | - | - | - |
-| imagen-4.0-ultra-generate-001 | vertexai | - | - | - |
-| imagetext | vertexai | - | - | - |
-| multimodalembedding | vertexai | - | - | - |
-| text-bison | vertexai | - | - | - |
| text-embedding-004 | vertexai | - | - | - |
| text-embedding-005 | vertexai | - | - | - |
| text-multilingual-embedding-002 | vertexai | - | - | - |
-| text-unicorn | vertexai | - | - | - |
-| textembedding-gecko | vertexai | - | - | - |
+
+
+### XAI (9)
+
+| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
+| :-- | :-- | --: | --: | :-- |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-non-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
## Models by Capability
-### Function Calling (356)
+### Function Calling (387)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
-| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
+| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00 |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| claude-opus-4-20250514 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| claude-opus-4-1-20250805 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
-| claude-3-7-sonnet-20250219 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| claude-sonnet-4-20250514 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| claude-3-7-sonnet-20250219 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
@@ -740,14 +747,14 @@ redirect_from:
| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| us.anthropic.claude-opus-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
@@ -755,42 +762,45 @@ redirect_from:
| us.anthropic.claude-sonnet-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| deepseek-chat | deepseek | 128000 | 8000 | In: $0.56, Out: $1.68, Cache: $0.07 |
| deepseek-reasoner | deepseek | - | 64000 | - |
+| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
-| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
+| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
+| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-latest | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
-| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-live-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-pro-exp | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp-02-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
-| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-live-2.5-flash-preview | gemini | 1048576 | 8192 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-exp-native-audio-thinking-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.5-flash-preview-native-audio-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.5-flash-native-audio-preview-09-2025 | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.5-flash-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-pro | gemini | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| gemini-2.5-pro-preview-06-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-03-25 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-05-06 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
-| gemini-embedding-exp | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-exp-1206 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-pro-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-robotics-er-1.5-preview | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
| codestral-2411-rc5 | mistral | 32768 | 8192 | - |
| codestral-2412 | mistral | 32768 | 8192 | - |
@@ -818,14 +828,14 @@ redirect_from:
| mistral-large-2411 | mistral | 32768 | 8192 | - |
| mistral-large-latest | mistral | 32768 | 8192 | - |
| mistral-large-pixtral-2411 | mistral | 32768 | 8192 | - |
-| mistral-medium | mistral | 32768 | 8192 | - |
| mistral-medium-2505 | mistral | 32768 | 8192 | - |
+| mistral-medium | mistral | 32768 | 8192 | - |
| mistral-medium-2508 | mistral | 32768 | 8192 | - |
| mistral-medium-latest | mistral | 32768 | 8192 | - |
| mistral-saba-2502 | mistral | 32768 | 8192 | - |
| mistral-saba-latest | mistral | 32768 | 8192 | - |
-| mistral-small | mistral | 32768 | 8192 | - |
| mistral-small-2409 | mistral | 32768 | 8192 | - |
+| mistral-small | mistral | 32768 | 8192 | - |
| mistral-small-2501 | mistral | 32768 | 8192 | - |
| mistral-small-2503 | mistral | 32768 | 8192 | - |
| mistral-small-2506 | mistral | 32768 | 8192 | - |
@@ -844,35 +854,57 @@ redirect_from:
| pixtral-12b-latest | mistral | 32768 | 8192 | - |
| pixtral-large-2411 | mistral | 32768 | 8192 | - |
| pixtral-large-latest | mistral | 32768 | 8192 | - |
-| gpt-4 | openai | 8192 | 8192 | In: $10.00, Out: $30.00 |
| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4-turbo-preview | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
-| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-5 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-chat-latest | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| gpt-4o-audio-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-realtime-preview | openai | 16000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 16000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-codex | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| codex-mini-latest | openai | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
+| computer-use-preview-2025-03-11 | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| gpt-audio | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-audio-2025-08-28 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-oss-120b | openai | 131072 | 131072 | - |
+| gpt-oss-20b | openai | 131072 | 131072 | - |
+| gpt-realtime | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
+| gpt-realtime-2025-08-28 | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
+| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-preview | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-preview-2024-09-12 | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
-| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40 |
-| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40 |
+| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-pro | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o3-pro-2025-06-10 | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o4-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini-2025-04-16 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
| ai21/jamba-large-1.7 | openrouter | 256000 | 4096 | In: $2.00, Out: $8.00 |
| ai21/jamba-mini-1.7 | openrouter | 256000 | 4096 | In: $0.20, Out: $0.40 |
| amazon/nova-lite-v1 | openrouter | 300000 | 5120 | In: $0.06, Out: $0.24 |
@@ -891,29 +923,26 @@ redirect_from:
| anthropic/claude-sonnet-4 | openrouter | 1000000 | 64000 | In: $3.00, Out: $15.00, Cache: $0.30 |
| arcee-ai/virtuoso-large | openrouter | 131072 | 64000 | In: $0.75, Out: $1.20 |
| deepcogito/cogito-v2-preview-llama-109b-moe | openrouter | 32767 | - | In: $0.18, Out: $0.59 |
-| cohere/command-r | openrouter | 128000 | 4000 | In: $0.50, Out: $1.50 |
-| cohere/command-r-03-2024 | openrouter | 128000 | 4000 | In: $0.50, Out: $1.50 |
| cohere/command-r-08-2024 | openrouter | 128000 | 4000 | In: $0.15, Out: $0.60 |
-| cohere/command-r-plus | openrouter | 128000 | 4000 | In: $3.00, Out: $15.00 |
-| cohere/command-r-plus-04-2024 | openrouter | 128000 | 4000 | In: $3.00, Out: $15.00 |
| cohere/command-r-plus-08-2024 | openrouter | 128000 | 4000 | In: $2.50, Out: $10.00 |
| deepseek/deepseek-chat | openrouter | 163840 | - | In: $0.25, Out: $1.00 |
| deepseek/deepseek-chat-v3-0324 | openrouter | 163840 | - | In: $0.25, Out: $1.00 |
| deepseek/deepseek-chat-v3-0324:free | openrouter | 163840 | - | - |
| deepseek/deepseek-chat-v3.1 | openrouter | 163840 | - | In: $0.25, Out: $1.00 |
| deepseek/deepseek-chat-v3.1:free | openrouter | 163840 | - | - |
+| deepseek/deepseek-v3.1-terminus | openrouter | 163840 | - | In: $0.27, Out: $1.00 |
| deepseek/deepseek-r1 | openrouter | 163840 | 163840 | In: $0.40, Out: $2.00 |
| deepseek/deepseek-r1-0528 | openrouter | 163840 | - | In: $0.40, Out: $1.75 |
| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.13 |
-| google/gemini-flash-1.5 | openrouter | 1000000 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| google/gemini-flash-1.5-8b | openrouter | 1000000 | 8192 | In: $0.04, Out: $0.15, Cache: $0.01 |
-| google/gemini-pro-1.5 | openrouter | 2000000 | 8192 | In: $1.25, Out: $5.00 |
| google/gemini-2.0-flash-001 | openrouter | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.0-flash-exp:free | openrouter | 1048576 | 8192 | - |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| google/gemini-2.5-flash-lite-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| google/gemini-2.5-flash-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview-05-06 | openrouter | 1048576 | 65535 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -926,7 +955,7 @@ redirect_from:
| meta-llama/llama-3.1-70b-instruct | openrouter | 131072 | 16384 | In: $0.10, Out: $0.28 |
| meta-llama/llama-3.1-8b-instruct | openrouter | 16384 | 16384 | In: $0.02, Out: $0.03 |
| meta-llama/llama-3.2-3b-instruct | openrouter | 16384 | 16384 | In: $0.02, Out: $0.02 |
-| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 131072 | In: $0.01, Out: $0.04 |
+| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 131072 | In: $0.04, Out: $0.12 |
| meta-llama/llama-3.3-70b-instruct:free | openrouter | 65536 | - | - |
| meta-llama/llama-3.3-8b-instruct:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-maverick | openrouter | 1048576 | 16384 | In: $0.15, Out: $0.60 |
@@ -971,7 +1000,6 @@ redirect_from:
| mistralai/pixtral-large-2411 | openrouter | 131072 | - | In: $2.00, Out: $6.00 |
| mistralai/mistral-saba | openrouter | 32768 | - | In: $0.20, Out: $0.60 |
| moonshotai/kimi-k2 | openrouter | 63000 | 63000 | In: $0.14, Out: $2.49 |
-| moonshotai/kimi-k2:free | openrouter | 32768 | - | - |
| moonshotai/kimi-k2-0905 | openrouter | 262144 | - | In: $0.38, Out: $1.52 |
| nvidia/llama-3.1-nemotron-70b-instruct | openrouter | 131072 | 16384 | In: $0.60, Out: $0.60 |
| nvidia/nemotron-nano-9b-v2 | openrouter | 131072 | - | In: $0.04, Out: $0.16 |
@@ -999,7 +1027,8 @@ redirect_from:
| openai/gpt-4o-audio-preview | openrouter | 128000 | 16384 | In: $2.50, Out: $10.00 |
| openai/gpt-4o-mini | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| openai/gpt-4o-mini-2024-07-18 | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
-| openai/gpt-5 | openrouter | 400000 | 128000 | In: $0.62, Out: $5.00, Cache: $0.06 |
+| openai/gpt-5 | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| openai/gpt-5-codex | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| openai/gpt-5-mini | openrouter | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| openai/gpt-5-nano | openrouter | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| openai/gpt-oss-120b | openrouter | 131072 | - | In: $0.05, Out: $0.25 |
@@ -1018,10 +1047,10 @@ redirect_from:
| qwen/qwen-max | openrouter | 32768 | 8192 | In: $1.60, Out: $6.40, Cache: $0.64 |
| qwen/qwen-plus | openrouter | 131072 | 8192 | In: $0.40, Out: $1.20, Cache: $0.16 |
| qwen/qwen-turbo | openrouter | 1000000 | 8192 | In: $0.05, Out: $0.20, Cache: $0.02 |
-| qwen/qwen3-14b | openrouter | 40960 | 40960 | In: $0.06, Out: $0.24 |
+| qwen/qwen3-14b | openrouter | 40960 | - | In: $0.04, Out: $0.14 |
| qwen/qwen3-235b-a22b | openrouter | 40960 | 40960 | In: $0.18, Out: $0.54 |
| qwen/qwen3-235b-a22b:free | openrouter | 131072 | - | - |
-| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | 262144 | In: $0.10, Out: $0.10 |
+| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | - | In: $0.09, Out: $0.60 |
| qwen/qwen3-235b-a22b-thinking-2507 | openrouter | 262144 | - | In: $0.10, Out: $0.39 |
| qwen/qwen3-30b-a3b | openrouter | 40960 | - | In: $0.06, Out: $0.22 |
| qwen/qwen3-30b-a3b-instruct-2507 | openrouter | 262144 | - | In: $0.07, Out: $0.28 |
@@ -1036,12 +1065,14 @@ redirect_from:
| qwen/qwen3-max | openrouter | 256000 | 32768 | In: $1.20, Out: $6.00, Cache: $0.24 |
| qwen/qwen3-next-80b-a3b-instruct | openrouter | 262144 | - | In: $0.10, Out: $0.80 |
| qwen/qwen3-next-80b-a3b-thinking | openrouter | 262144 | - | In: $0.10, Out: $0.80 |
+| qwen/qwen3-vl-235b-a22b-instruct | openrouter | 131072 | 32768 | In: $0.30, Out: $1.50 |
+| qwen/qwen3-vl-235b-a22b-thinking | openrouter | 131072 | 32768 | In: $0.30, Out: $3.00 |
| stepfun-ai/step3 | openrouter | 65536 | 65536 | In: $0.57, Out: $1.42 |
| thedrummer/rocinante-12b | openrouter | 32768 | - | In: $0.17, Out: $0.43 |
| thedrummer/unslopnemo-12b | openrouter | 32768 | - | In: $0.40, Out: $0.40 |
| alibaba/tongyi-deepresearch-30b-a3b | openrouter | 131072 | 131072 | In: $0.09, Out: $0.45 |
| z-ai/glm-4-32b | openrouter | 128000 | - | In: $0.10, Out: $0.10 |
-| z-ai/glm-4.5 | openrouter | 131072 | - | In: $0.41, Out: $1.65 |
+| z-ai/glm-4.5 | openrouter | 131072 | - | In: $0.40, Out: $1.60 |
| z-ai/glm-4.5-air | openrouter | 131072 | 131072 | In: $0.14, Out: $0.86 |
| z-ai/glm-4.5-air:free | openrouter | 131072 | - | - |
| z-ai/glm-4.5v | openrouter | 65536 | 65536 | In: $0.50, Out: $1.80 |
@@ -1050,6 +1081,7 @@ redirect_from:
| x-ai/grok-3-mini | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-3-mini-beta | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| x-ai/grok-4-fast | openrouter | 2000000 | 30000 | In: $0.20, Out: $0.50, Cache: $0.05 |
| x-ai/grok-4-fast:free | openrouter | 2000000 | 30000 | - |
| x-ai/grok-code-fast-1 | openrouter | 256000 | 10000 | In: $0.20, Out: $1.50, Cache: $0.02 |
| gemini-1.5-flash | vertexai | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
@@ -1072,24 +1104,33 @@ redirect_from:
| text-embedding-004 | vertexai | - | - | - |
| text-embedding-005 | vertexai | - | - | - |
| text-multilingual-embedding-002 | vertexai | - | - | - |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-non-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
-### Structured Output (335)
+### Structured Output (363)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
| deepseek-chat | deepseek | 128000 | 8000 | In: $0.56, Out: $1.68, Cache: $0.07 |
| deepseek-reasoner | deepseek | - | 64000 | - |
-| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
-| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-latest | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
@@ -1104,19 +1145,24 @@ redirect_from:
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-image-preview | gemini | 32768 | 32768 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-live-2.5-flash-preview | gemini | 1048576 | 8192 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
-| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
-| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-preview-06-17 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-2.5-flash-lite-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro | gemini | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| gemini-2.5-pro-preview-06-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-03-25 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-05-06 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
-| gemini-embedding-exp | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-exp-1206 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-lite-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-pro-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-robotics-er-1.5-preview | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
| codestral-2411-rc5 | mistral | 32768 | 8192 | - |
| codestral-2412 | mistral | 32768 | 8192 | - |
@@ -1144,16 +1190,16 @@ redirect_from:
| mistral-large-2411 | mistral | 32768 | 8192 | - |
| mistral-large-latest | mistral | 32768 | 8192 | - |
| mistral-large-pixtral-2411 | mistral | 32768 | 8192 | - |
-| mistral-medium | mistral | 32768 | 8192 | - |
| mistral-medium-2505 | mistral | 32768 | 8192 | - |
| mistral-medium-2508 | mistral | 32768 | 8192 | - |
| mistral-medium-latest | mistral | 32768 | 8192 | - |
+| mistral-medium | mistral | 32768 | 8192 | - |
| mistral-saba-2502 | mistral | 32768 | 8192 | - |
| mistral-saba-latest | mistral | 32768 | 8192 | - |
+| mistral-small-2503 | mistral | 32768 | 8192 | - |
| mistral-small | mistral | 32768 | 8192 | - |
| mistral-small-2409 | mistral | 32768 | 8192 | - |
| mistral-small-2501 | mistral | 32768 | 8192 | - |
-| mistral-small-2503 | mistral | 32768 | 8192 | - |
| mistral-small-2506 | mistral | 32768 | 8192 | - |
| mistral-small-latest | mistral | 32768 | 8192 | - |
| mistral-tiny | mistral | 32768 | 8192 | - |
@@ -1165,48 +1211,59 @@ redirect_from:
| open-mixtral-8x22b | mistral | 32768 | 8192 | - |
| open-mixtral-8x22b-2404 | mistral | 32768 | 8192 | - |
| open-mixtral-8x7b | mistral | 32768 | 8192 | - |
-| pixtral-12b | mistral | 32768 | 8192 | - |
| pixtral-12b-2409 | mistral | 32768 | 8192 | - |
| pixtral-12b-latest | mistral | 32768 | 8192 | - |
+| pixtral-12b | mistral | 32768 | 8192 | - |
| pixtral-large-2411 | mistral | 32768 | 8192 | - |
| pixtral-large-latest | mistral | 32768 | 8192 | - |
-| chatgpt-4o-latest | openai | 128000 | 16384 | In: $5.00, Out: $15.00 |
-| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
-| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-5 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-chat-latest | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
+| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini-search-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-codex | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| codex-mini-latest | openai | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
+| gpt-oss-120b | openai | 131072 | 131072 | - |
+| gpt-oss-20b | openai | 131072 | 131072 | - |
+| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-preview-2024-09-12 | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-preview | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
-| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40 |
-| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40 |
+| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-pro | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o3-pro-2025-06-10 | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o4-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini-2025-04-16 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
| ai21/jamba-large-1.7 | openrouter | 256000 | 4096 | In: $2.00, Out: $8.00 |
| ai21/jamba-mini-1.7 | openrouter | 256000 | 4096 | In: $0.20, Out: $0.40 |
| arcee-ai/afm-4.5b | openrouter | 65536 | - | In: $0.10, Out: $0.40 |
| baidu/ernie-4.5-300b-a47b | openrouter | 123000 | 12000 | In: $0.28, Out: $1.10 |
-| cohere/command | openrouter | 4096 | 4000 | In: $1.00, Out: $2.00 |
| cohere/command-a | openrouter | 256000 | 8192 | In: $2.50, Out: $10.00 |
-| cohere/command-r | openrouter | 128000 | 4000 | In: $0.50, Out: $1.50 |
-| cohere/command-r-03-2024 | openrouter | 128000 | 4000 | In: $0.50, Out: $1.50 |
| cohere/command-r-08-2024 | openrouter | 128000 | 4000 | In: $0.15, Out: $0.60 |
-| cohere/command-r-plus | openrouter | 128000 | 4000 | In: $3.00, Out: $15.00 |
-| cohere/command-r-plus-04-2024 | openrouter | 128000 | 4000 | In: $3.00, Out: $15.00 |
| cohere/command-r-plus-08-2024 | openrouter | 128000 | 4000 | In: $2.50, Out: $10.00 |
| cohere/command-r7b-12-2024 | openrouter | 128000 | 4000 | In: $0.04, Out: $0.15 |
| deepseek/deepseek-prover-v2 | openrouter | 163840 | - | In: $0.50, Out: $2.18 |
@@ -1214,21 +1271,22 @@ redirect_from:
| deepseek/deepseek-chat-v3-0324 | openrouter | 163840 | - | In: $0.25, Out: $1.00 |
| deepseek/deepseek-chat-v3.1 | openrouter | 163840 | - | In: $0.25, Out: $1.00 |
| deepseek/deepseek-chat-v3.1:free | openrouter | 163840 | - | - |
+| deepseek/deepseek-v3.1-terminus | openrouter | 163840 | - | In: $0.27, Out: $1.00 |
| deepseek/deepseek-r1 | openrouter | 163840 | 163840 | In: $0.40, Out: $2.00 |
| deepseek/deepseek-r1-0528 | openrouter | 163840 | - | In: $0.40, Out: $1.75 |
| deepseek/deepseek-r1-distill-llama-70b | openrouter | 131072 | - | In: $0.03, Out: $0.13 |
| deepseek/deepseek-r1-distill-qwen-32b | openrouter | 131072 | 16384 | In: $0.27, Out: $0.27 |
+| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
| alpindale/goliath-120b | openrouter | 6144 | 512 | In: $4.00, Out: $5.50 |
-| google/gemini-flash-1.5 | openrouter | 1000000 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| google/gemini-flash-1.5-8b | openrouter | 1000000 | 8192 | In: $0.04, Out: $0.15, Cache: $0.01 |
-| google/gemini-pro-1.5 | openrouter | 2000000 | 8192 | In: $1.25, Out: $5.00 |
| google/gemini-2.0-flash-001 | openrouter | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.0-flash-exp:free | openrouter | 1048576 | 8192 | - |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| google/gemini-2.5-flash-lite-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| google/gemini-2.5-flash-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview-05-06 | openrouter | 1048576 | 65535 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -1248,14 +1306,13 @@ redirect_from:
| meta-llama/llama-3-70b-instruct | openrouter | 8192 | 16384 | In: $0.30, Out: $0.40 |
| meta-llama/llama-3-8b-instruct | openrouter | 8192 | 16384 | In: $0.03, Out: $0.06 |
| meta-llama/llama-3.1-405b-instruct | openrouter | 32768 | 16384 | In: $0.80, Out: $0.80 |
-| meta-llama/llama-3.1-405b-instruct:free | openrouter | 65536 | - | - |
| meta-llama/llama-3.1-70b-instruct | openrouter | 131072 | 16384 | In: $0.10, Out: $0.28 |
| meta-llama/llama-3.1-8b-instruct | openrouter | 16384 | 16384 | In: $0.02, Out: $0.03 |
| meta-llama/llama-3.2-11b-vision-instruct | openrouter | 131072 | 16384 | In: $0.05, Out: $0.05 |
| meta-llama/llama-3.2-1b-instruct | openrouter | 131072 | 16384 | In: $0.01, Out: $0.01 |
| meta-llama/llama-3.2-3b-instruct | openrouter | 16384 | 16384 | In: $0.02, Out: $0.02 |
| meta-llama/llama-3.2-90b-vision-instruct | openrouter | 32768 | 16384 | In: $0.35, Out: $0.40 |
-| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 131072 | In: $0.01, Out: $0.04 |
+| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 131072 | In: $0.04, Out: $0.12 |
| meta-llama/llama-3.3-8b-instruct:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-maverick | openrouter | 1048576 | 16384 | In: $0.15, Out: $0.60 |
| meta-llama/llama-4-maverick:free | openrouter | 128000 | 4028 | - |
@@ -1331,8 +1388,9 @@ redirect_from:
| openai/gpt-4o-mini | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| openai/gpt-4o-mini-2024-07-18 | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| openai/gpt-4o-mini-search-preview | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| openai/gpt-5 | openrouter | 400000 | 128000 | In: $0.62, Out: $5.00, Cache: $0.06 |
+| openai/gpt-5 | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| openai/gpt-5-chat | openrouter | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| openai/gpt-5-codex | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| openai/gpt-5-mini | openrouter | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| openai/gpt-5-nano | openrouter | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| openai/gpt-oss-120b | openrouter | 131072 | - | In: $0.05, Out: $0.25 |
@@ -1351,7 +1409,6 @@ redirect_from:
| qwen/qwen-2.5-7b-instruct | openrouter | 65536 | - | In: $0.04, Out: $0.10 |
| qwen/qwen-2.5-coder-32b-instruct | openrouter | 32768 | 16384 | In: $0.06, Out: $0.15 |
| qwen/qwq-32b | openrouter | 32768 | - | In: $0.15, Out: $0.40 |
-| qwen/qwq-32b:free | openrouter | 32768 | - | - |
| qwen/qwen-plus-2025-07-28 | openrouter | 1000000 | 32768 | In: $0.40, Out: $1.20 |
| qwen/qwen-plus-2025-07-28:thinking | openrouter | 1000000 | 32768 | In: $0.40, Out: $4.00 |
| qwen/qwen-vl-max | openrouter | 7500 | 1500 | In: $0.80, Out: $3.20 |
@@ -1361,12 +1418,12 @@ redirect_from:
| qwen/qwen-turbo | openrouter | 1000000 | 8192 | In: $0.05, Out: $0.20, Cache: $0.02 |
| qwen/qwen2.5-vl-32b-instruct | openrouter | 16384 | - | In: $0.04, Out: $0.14 |
| qwen/qwen2.5-vl-32b-instruct:free | openrouter | 8192 | - | - |
-| qwen/qwen2.5-vl-72b-instruct:free | openrouter | 32768 | - | - |
+| qwen/qwen2.5-vl-72b-instruct:free | openrouter | 131072 | 2048 | - |
| qwen/qwen-2.5-vl-7b-instruct | openrouter | 32768 | - | In: $0.20, Out: $0.20 |
-| qwen/qwen3-14b | openrouter | 40960 | 40960 | In: $0.06, Out: $0.24 |
+| qwen/qwen3-14b | openrouter | 40960 | - | In: $0.04, Out: $0.14 |
| qwen/qwen3-235b-a22b | openrouter | 40960 | 40960 | In: $0.18, Out: $0.54 |
| qwen/qwen3-235b-a22b:free | openrouter | 131072 | - | - |
-| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | 262144 | In: $0.10, Out: $0.10 |
+| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | - | In: $0.09, Out: $0.60 |
| qwen/qwen3-235b-a22b-thinking-2507 | openrouter | 262144 | - | In: $0.10, Out: $0.39 |
| qwen/qwen3-30b-a3b | openrouter | 40960 | - | In: $0.06, Out: $0.22 |
| qwen/qwen3-30b-a3b-instruct-2507 | openrouter | 262144 | - | In: $0.07, Out: $0.28 |
@@ -1380,25 +1437,28 @@ redirect_from:
| qwen/qwen3-max | openrouter | 256000 | 32768 | In: $1.20, Out: $6.00, Cache: $0.24 |
| qwen/qwen3-next-80b-a3b-instruct | openrouter | 262144 | - | In: $0.10, Out: $0.80 |
| qwen/qwen3-next-80b-a3b-thinking | openrouter | 262144 | - | In: $0.10, Out: $0.80 |
+| qwen/qwen3-vl-235b-a22b-instruct | openrouter | 131072 | 32768 | In: $0.30, Out: $1.50 |
+| qwen/qwen3-vl-235b-a22b-thinking | openrouter | 131072 | 32768 | In: $0.30, Out: $3.00 |
| undi95/remm-slerp-l2-13b | openrouter | 6144 | - | In: $0.45, Out: $0.65 |
| sao10k/l3-lunaris-8b | openrouter | 8192 | - | In: $0.04, Out: $0.05 |
| sao10k/l3.1-euryale-70b | openrouter | 32768 | - | In: $0.65, Out: $0.75 |
| sao10k/l3.3-euryale-70b | openrouter | 131072 | 16384 | In: $0.65, Out: $0.75 |
| stepfun-ai/step3 | openrouter | 65536 | 65536 | In: $0.57, Out: $1.42 |
| tencent/hunyuan-a13b-instruct | openrouter | 32768 | - | In: $0.03, Out: $0.03 |
-| thedrummer/anubis-70b-v1.1 | openrouter | 16384 | - | In: $0.40, Out: $0.70 |
+| thedrummer/anubis-70b-v1.1 | openrouter | 4096 | - | In: $0.40, Out: $0.70 |
| thedrummer/rocinante-12b | openrouter | 32768 | - | In: $0.17, Out: $0.43 |
| thedrummer/unslopnemo-12b | openrouter | 32768 | - | In: $0.40, Out: $0.40 |
| alibaba/tongyi-deepresearch-30b-a3b | openrouter | 131072 | 131072 | In: $0.09, Out: $0.45 |
| cognitivecomputations/dolphin-mistral-24b-venice-edition:free | openrouter | 32768 | - | - |
| microsoft/wizardlm-2-8x22b | openrouter | 65536 | 65536 | In: $0.48, Out: $0.48 |
-| z-ai/glm-4.5 | openrouter | 131072 | - | In: $0.41, Out: $1.65 |
+| z-ai/glm-4.5 | openrouter | 131072 | - | In: $0.40, Out: $1.60 |
| z-ai/glm-4.5-air | openrouter | 131072 | 131072 | In: $0.14, Out: $0.86 |
| x-ai/grok-3 | openrouter | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
| x-ai/grok-3-beta | openrouter | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
| x-ai/grok-3-mini | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-3-mini-beta | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| x-ai/grok-4-fast | openrouter | 2000000 | 30000 | In: $0.20, Out: $0.50, Cache: $0.05 |
| x-ai/grok-4-fast:free | openrouter | 2000000 | 30000 | - |
| x-ai/grok-code-fast-1 | openrouter | 256000 | 10000 | In: $0.20, Out: $1.50, Cache: $0.02 |
| gemini-1.5-flash | vertexai | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
@@ -1413,38 +1473,53 @@ redirect_from:
| gemini-2.5-flash | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-pro | vertexai | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-non-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
-### Streaming (557)
+### Streaming (489)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
+| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00 |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| claude-opus-4-20250514 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| claude-opus-4-1-20250805 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
-| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| claude-3-7-sonnet-20250219 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
+| claude-sonnet-4-20250514 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| us.anthropic.claude-opus-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
@@ -1459,25 +1534,25 @@ redirect_from:
| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.5-flash-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash-lite-preview-06-17 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-2.5-flash-lite-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-06-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-03-25 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-05-06 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
-| gemini-embedding-001 | gemini | 2048 | 1 | - |
-| gemini-embedding-exp | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-exp-1206 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-lite-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-pro-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-robotics-er-1.5-preview | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemma-3-12b-it | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemma-3-1b-it | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemma-3-27b-it | gemini | 131072 | 8192 | In: $0.08, Out: $0.30 |
| gemma-3-4b-it | gemini | 32768 | 8192 | In: $0.08, Out: $0.30 |
| gemma-3n-e2b-it | gemini | 8192 | 2048 | In: $0.08, Out: $0.30 |
| gemma-3n-e4b-it | gemini | 8192 | 2048 | In: $0.08, Out: $0.30 |
-| imagen-3.0-generate-002 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-preview-06-06 | gemini | 480 | 8192 | - |
-| imagen-4.0-ultra-generate-preview-06-06 | gemini | 480 | 8192 | - |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
-| aqa | gemini | 7168 | 1024 | - |
| text-embedding-004 | gemini | 2048 | 1 | - |
| codestral-2411-rc5 | mistral | 32768 | 8192 | - |
| codestral-2412 | mistral | 32768 | 8192 | - |
@@ -1537,91 +1612,24 @@ redirect_from:
| voxtral-mini-latest | mistral | 32768 | 8192 | - |
| voxtral-small-2507 | mistral | 32768 | 8192 | - |
| voxtral-small-latest | mistral | 32768 | 8192 | - |
-| babbage-002 | openai | 4096 | 16384 | In: $0.40, Out: $0.40 |
-| chatgpt-4o-latest | openai | 128000 | 16384 | In: $5.00, Out: $15.00 |
-| codex-mini-latest | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| dall-e-2 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| dall-e-3 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| davinci-002 | openai | 4096 | 16384 | In: $2.00, Out: $2.00 |
-| gpt-3.5-turbo | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-0125 | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-1106 | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-16k | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-instruct | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
| gpt-3.5-turbo-instruct-0914 | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
-| gpt-4 | openai | 8192 | 8192 | In: $10.00, Out: $30.00 |
| gpt-4-0125-preview | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-4-0613 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
| gpt-4-1106-preview | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4-turbo-preview | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
-| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
-| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-audio-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2024-10-01 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2025-06-03 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-realtime-preview | openai | 128000 | 4096 | In: $0.60, Out: $2.40 |
-| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $0.60, Out: $2.40 |
-| gpt-4o-mini-search-preview | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-4o-mini-search-preview-2025-03-11 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-4o-mini-transcribe | openai | 16000 | 2000 | In: $1.25, Out: $5.00 |
-| gpt-4o-mini-tts | openai | - | - | In: $0.60, Out: $12.00 |
| gpt-4o-realtime-preview | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2024-10-01 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2024-12-17 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
| gpt-4o-realtime-preview-2025-06-03 | openai | 128000 | 4096 | In: $5.00, Out: $20.00 |
-| gpt-4o-transcribe | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-5 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-chat-latest | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-audio | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-audio-2025-08-28 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-image-1 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-realtime | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-realtime-2025-08-28 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-mini | openai | 128000 | 65536 | In: $1.10, Out: $4.40 |
-| o1-mini-2024-09-12 | openai | 128000 | 65536 | In: $1.10, Out: $4.40 |
-| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
-| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
-| o3 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-2025-04-16 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-deep-research | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-deep-research-2025-06-26 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40 |
-| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40 |
-| o3-pro | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o3-pro-2025-06-10 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o4-mini | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o4-mini-2025-04-16 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o4-mini-deep-research | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| o4-mini-deep-research-2025-06-26 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| tts-1 | openai | - | - | In: $15.00, Out: $15.00 |
| tts-1-1106 | openai | - | - | In: $15.00, Out: $15.00 |
-| tts-1-hd | openai | - | - | In: $30.00, Out: $30.00 |
| tts-1-hd-1106 | openai | - | - | In: $30.00, Out: $30.00 |
-| whisper-1 | openai | - | - | In: $0.01, Out: $0.01 |
| ai21/jamba-large-1.7 | openrouter | 256000 | 4096 | In: $2.00, Out: $8.00 |
| ai21/jamba-mini-1.7 | openrouter | 256000 | 4096 | In: $0.20, Out: $0.40 |
| agentica-org/deepcoder-14b-preview | openrouter | 96000 | - | In: $0.02, Out: $0.02 |
@@ -1661,13 +1669,8 @@ redirect_from:
| bytedance/seed-oss-36b-instruct | openrouter | 131072 | - | In: $0.16, Out: $0.65 |
| bytedance/ui-tars-1.5-7b | openrouter | 128000 | 2048 | In: $0.10, Out: $0.20 |
| deepcogito/cogito-v2-preview-llama-109b-moe | openrouter | 32767 | - | In: $0.18, Out: $0.59 |
-| cohere/command | openrouter | 4096 | 4000 | In: $1.00, Out: $2.00 |
| cohere/command-a | openrouter | 256000 | 8192 | In: $2.50, Out: $10.00 |
-| cohere/command-r | openrouter | 128000 | 4000 | In: $0.50, Out: $1.50 |
-| cohere/command-r-03-2024 | openrouter | 128000 | 4000 | In: $0.50, Out: $1.50 |
| cohere/command-r-08-2024 | openrouter | 128000 | 4000 | In: $0.15, Out: $0.60 |
-| cohere/command-r-plus | openrouter | 128000 | 4000 | In: $3.00, Out: $15.00 |
-| cohere/command-r-plus-04-2024 | openrouter | 128000 | 4000 | In: $3.00, Out: $15.00 |
| cohere/command-r-plus-08-2024 | openrouter | 128000 | 4000 | In: $2.50, Out: $10.00 |
| cohere/command-r7b-12-2024 | openrouter | 128000 | 4000 | In: $0.04, Out: $0.15 |
| deepcogito/cogito-v2-preview-deepseek-671b | openrouter | 163840 | - | In: $1.25, Out: $1.25 |
@@ -1678,6 +1681,7 @@ redirect_from:
| deepseek/deepseek-chat-v3.1 | openrouter | 163840 | - | In: $0.25, Out: $1.00 |
| deepseek/deepseek-chat-v3.1:free | openrouter | 163840 | - | - |
| deepseek/deepseek-v3.1-base | openrouter | 163840 | - | In: $0.25, Out: $1.00 |
+| deepseek/deepseek-v3.1-terminus | openrouter | 163840 | - | In: $0.27, Out: $1.00 |
| deepseek/deepseek-r1-0528-qwen3-8b | openrouter | 131072 | - | In: $0.01, Out: $0.05 |
| deepseek/deepseek-r1-0528-qwen3-8b:free | openrouter | 131072 | - | - |
| deepseek/deepseek-r1 | openrouter | 163840 | 163840 | In: $0.40, Out: $2.00 |
@@ -1694,17 +1698,17 @@ redirect_from:
| cognitivecomputations/dolphin3.0-r1-mistral-24b | openrouter | 32768 | - | In: $0.01, Out: $0.03 |
| cognitivecomputations/dolphin3.0-r1-mistral-24b:free | openrouter | 32768 | - | - |
| eleutherai/llemma_7b | openrouter | 4096 | 4096 | In: $0.80, Out: $1.20 |
+| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
| alpindale/goliath-120b | openrouter | 6144 | 512 | In: $4.00, Out: $5.50 |
-| google/gemini-flash-1.5 | openrouter | 1000000 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| google/gemini-flash-1.5-8b | openrouter | 1000000 | 8192 | In: $0.04, Out: $0.15, Cache: $0.01 |
-| google/gemini-pro-1.5 | openrouter | 2000000 | 8192 | In: $1.25, Out: $5.00 |
| google/gemini-2.0-flash-001 | openrouter | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.0-flash-exp:free | openrouter | 1048576 | 8192 | - |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| google/gemini-2.5-flash-lite-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| google/gemini-2.5-flash-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview-05-06 | openrouter | 1048576 | 65535 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -1728,14 +1732,13 @@ redirect_from:
| liquid/lfm-7b | openrouter | 32768 | - | In: $0.01, Out: $0.01 |
| meta-llama/llama-guard-3-8b | openrouter | 131072 | - | In: $0.02, Out: $0.06 |
| anthracite-org/magnum-v2-72b | openrouter | 32768 | - | In: $3.00, Out: $3.00 |
-| anthracite-org/magnum-v4-72b | openrouter | 16384 | 2048 | In: $2.00, Out: $5.00 |
+| anthracite-org/magnum-v4-72b | openrouter | 16384 | 2048 | In: $2.50, Out: $5.00 |
| mancer/weaver | openrouter | 8000 | 2000 | In: $1.12, Out: $1.12 |
| meituan/longcat-flash-chat | openrouter | 131072 | - | In: $0.12, Out: $0.60 |
| meta-llama/llama-3-70b-instruct | openrouter | 8192 | 16384 | In: $0.30, Out: $0.40 |
| meta-llama/llama-3-8b-instruct | openrouter | 8192 | 16384 | In: $0.03, Out: $0.06 |
| meta-llama/llama-3.1-405b | openrouter | 32768 | - | In: $2.00, Out: $2.00 |
| meta-llama/llama-3.1-405b-instruct | openrouter | 32768 | 16384 | In: $0.80, Out: $0.80 |
-| meta-llama/llama-3.1-405b-instruct:free | openrouter | 65536 | - | - |
| meta-llama/llama-3.1-70b-instruct | openrouter | 131072 | 16384 | In: $0.10, Out: $0.28 |
| meta-llama/llama-3.1-8b-instruct | openrouter | 16384 | 16384 | In: $0.02, Out: $0.03 |
| meta-llama/llama-3.2-11b-vision-instruct | openrouter | 131072 | 16384 | In: $0.05, Out: $0.05 |
@@ -1743,7 +1746,7 @@ redirect_from:
| meta-llama/llama-3.2-3b-instruct | openrouter | 16384 | 16384 | In: $0.02, Out: $0.02 |
| meta-llama/llama-3.2-3b-instruct:free | openrouter | 131072 | - | - |
| meta-llama/llama-3.2-90b-vision-instruct | openrouter | 32768 | 16384 | In: $0.35, Out: $0.40 |
-| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 131072 | In: $0.01, Out: $0.04 |
+| meta-llama/llama-3.3-70b-instruct | openrouter | 131072 | 131072 | In: $0.04, Out: $0.12 |
| meta-llama/llama-3.3-70b-instruct:free | openrouter | 65536 | - | - |
| meta-llama/llama-3.3-8b-instruct:free | openrouter | 128000 | 4028 | - |
| meta-llama/llama-4-maverick | openrouter | 1048576 | 16384 | In: $0.15, Out: $0.60 |
@@ -1814,6 +1817,7 @@ redirect_from:
| neversleep/llama-3-lumimaid-70b | openrouter | 8192 | 4096 | In: $4.00, Out: $6.00 |
| neversleep/llama-3.1-lumimaid-8b | openrouter | 32768 | - | In: $0.09, Out: $0.60 |
| neversleep/noromaid-20b | openrouter | 4096 | - | In: $1.00, Out: $1.75 |
+| nousresearch/deephermes-3-llama-3-8b-preview | openrouter | 131072 | - | In: $0.01, Out: $0.05 |
| nousresearch/deephermes-3-llama-3-8b-preview:free | openrouter | 131072 | - | - |
| nousresearch/deephermes-3-mistral-24b-preview | openrouter | 32768 | - | In: $0.13, Out: $0.51 |
| nousresearch/hermes-3-llama-3.1-405b | openrouter | 131072 | 131072 | In: $0.80, Out: $0.80 |
@@ -1845,8 +1849,9 @@ redirect_from:
| openai/gpt-4o-mini | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| openai/gpt-4o-mini-2024-07-18 | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| openai/gpt-4o-mini-search-preview | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| openai/gpt-5 | openrouter | 400000 | 128000 | In: $0.62, Out: $5.00, Cache: $0.06 |
+| openai/gpt-5 | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| openai/gpt-5-chat | openrouter | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| openai/gpt-5-codex | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| openai/gpt-5-mini | openrouter | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| openai/gpt-5-nano | openrouter | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| openai/gpt-oss-120b | openrouter | 131072 | - | In: $0.05, Out: $0.25 |
@@ -1876,8 +1881,6 @@ redirect_from:
| qwen/qwen-2.5-coder-32b-instruct | openrouter | 32768 | 16384 | In: $0.06, Out: $0.15 |
| qwen/qwen-2.5-coder-32b-instruct:free | openrouter | 32768 | - | - |
| qwen/qwq-32b | openrouter | 32768 | - | In: $0.15, Out: $0.40 |
-| qwen/qwq-32b:free | openrouter | 32768 | - | - |
-| qwen/qwq-32b-preview | openrouter | 32768 | - | In: $0.20, Out: $0.20 |
| qwen/qwen-plus-2025-07-28 | openrouter | 1000000 | 32768 | In: $0.40, Out: $1.20 |
| qwen/qwen-plus-2025-07-28:thinking | openrouter | 1000000 | 32768 | In: $0.40, Out: $4.00 |
| qwen/qwen-vl-max | openrouter | 7500 | 1500 | In: $0.80, Out: $3.20 |
@@ -1888,13 +1891,13 @@ redirect_from:
| qwen/qwen2.5-vl-32b-instruct | openrouter | 16384 | - | In: $0.04, Out: $0.14 |
| qwen/qwen2.5-vl-32b-instruct:free | openrouter | 8192 | - | - |
| qwen/qwen2.5-vl-72b-instruct | openrouter | 32768 | - | In: $0.07, Out: $0.28 |
-| qwen/qwen2.5-vl-72b-instruct:free | openrouter | 32768 | - | - |
+| qwen/qwen2.5-vl-72b-instruct:free | openrouter | 131072 | 2048 | - |
| qwen/qwen-2.5-vl-7b-instruct | openrouter | 32768 | - | In: $0.20, Out: $0.20 |
-| qwen/qwen3-14b | openrouter | 40960 | 40960 | In: $0.06, Out: $0.24 |
+| qwen/qwen3-14b | openrouter | 40960 | - | In: $0.04, Out: $0.14 |
| qwen/qwen3-14b:free | openrouter | 40960 | - | - |
| qwen/qwen3-235b-a22b | openrouter | 40960 | 40960 | In: $0.18, Out: $0.54 |
| qwen/qwen3-235b-a22b:free | openrouter | 131072 | - | - |
-| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | 262144 | In: $0.10, Out: $0.10 |
+| qwen/qwen3-235b-a22b-2507 | openrouter | 262144 | - | In: $0.09, Out: $0.60 |
| qwen/qwen3-235b-a22b-thinking-2507 | openrouter | 262144 | - | In: $0.10, Out: $0.39 |
| qwen/qwen3-30b-a3b | openrouter | 40960 | - | In: $0.06, Out: $0.22 |
| qwen/qwen3-30b-a3b:free | openrouter | 40960 | - | - |
@@ -1912,6 +1915,8 @@ redirect_from:
| qwen/qwen3-max | openrouter | 256000 | 32768 | In: $1.20, Out: $6.00, Cache: $0.24 |
| qwen/qwen3-next-80b-a3b-instruct | openrouter | 262144 | - | In: $0.10, Out: $0.80 |
| qwen/qwen3-next-80b-a3b-thinking | openrouter | 262144 | - | In: $0.10, Out: $0.80 |
+| qwen/qwen3-vl-235b-a22b-instruct | openrouter | 131072 | 32768 | In: $0.30, Out: $1.50 |
+| qwen/qwen3-vl-235b-a22b-thinking | openrouter | 131072 | 32768 | In: $0.30, Out: $3.00 |
| undi95/remm-slerp-l2-13b | openrouter | 6144 | - | In: $0.45, Out: $0.65 |
| sao10k/l3-lunaris-8b | openrouter | 8192 | - | In: $0.04, Out: $0.05 |
| sao10k/l3.1-euryale-70b | openrouter | 32768 | - | In: $0.65, Out: $0.75 |
@@ -1929,7 +1934,7 @@ redirect_from:
| tngtech/deepseek-r1t2-chimera:free | openrouter | 163840 | - | - |
| tencent/hunyuan-a13b-instruct | openrouter | 32768 | - | In: $0.03, Out: $0.03 |
| tencent/hunyuan-a13b-instruct:free | openrouter | 32768 | - | - |
-| thedrummer/anubis-70b-v1.1 | openrouter | 16384 | - | In: $0.40, Out: $0.70 |
+| thedrummer/anubis-70b-v1.1 | openrouter | 4096 | - | In: $0.40, Out: $0.70 |
| thedrummer/anubis-pro-105b-v1 | openrouter | 131072 | 131072 | In: $0.50, Out: $1.00 |
| thedrummer/rocinante-12b | openrouter | 32768 | - | In: $0.17, Out: $0.43 |
| thedrummer/skyfall-36b-v2 | openrouter | 32768 | - | In: $0.04, Out: $0.16 |
@@ -1938,7 +1943,7 @@ redirect_from:
| cognitivecomputations/dolphin-mistral-24b-venice-edition:free | openrouter | 32768 | - | - |
| microsoft/wizardlm-2-8x22b | openrouter | 65536 | 65536 | In: $0.48, Out: $0.48 |
| z-ai/glm-4-32b | openrouter | 128000 | - | In: $0.10, Out: $0.10 |
-| z-ai/glm-4.5 | openrouter | 131072 | - | In: $0.41, Out: $1.65 |
+| z-ai/glm-4.5 | openrouter | 131072 | - | In: $0.40, Out: $1.60 |
| z-ai/glm-4.5-air | openrouter | 131072 | 131072 | In: $0.14, Out: $0.86 |
| z-ai/glm-4.5-air:free | openrouter | 131072 | - | - |
| z-ai/glm-4.5v | openrouter | 65536 | 65536 | In: $0.50, Out: $1.80 |
@@ -1947,6 +1952,7 @@ redirect_from:
| x-ai/grok-3-mini | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-3-mini-beta | openrouter | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| x-ai/grok-4-fast | openrouter | 2000000 | 30000 | In: $0.20, Out: $0.50, Cache: $0.05 |
| x-ai/grok-4-fast:free | openrouter | 2000000 | 30000 | - |
| x-ai/grok-code-fast-1 | openrouter | 256000 | 10000 | In: $0.20, Out: $1.50, Cache: $0.02 |
| sonar | perplexity | 128000 | 4096 | In: $1.00, Out: $1.00 |
@@ -1954,43 +1960,48 @@ redirect_from:
| sonar-pro | perplexity | 200000 | 8192 | In: $3.00, Out: $15.00 |
| sonar-reasoning | perplexity | 128000 | 4096 | In: $1.00, Out: $5.00 |
| sonar-reasoning-pro | perplexity | 128000 | 8192 | In: $2.00, Out: $8.00 |
-| chat-bison | vertexai | - | - | - |
-| code-bison | vertexai | - | - | - |
-| code-gecko | vertexai | - | - | - |
-| codechat-bison | vertexai | - | - | - |
| gemini-embedding-001 | vertexai | - | - | - |
| gemini-exp-1121 | vertexai | - | - | - |
| gemini-exp-1206 | vertexai | - | - | - |
| gemini-pro | vertexai | - | - | - |
| gemini-pro-vision | vertexai | - | - | - |
-| image-segmentation-001 | vertexai | - | - | - |
-| imagegeneration | vertexai | - | - | - |
-| imagen-4.0-fast-generate-001 | vertexai | - | - | - |
-| imagen-4.0-generate-001 | vertexai | - | - | - |
-| imagen-4.0-ultra-generate-001 | vertexai | - | - | - |
-| imagetext | vertexai | - | - | - |
-| multimodalembedding | vertexai | - | - | - |
-| text-bison | vertexai | - | - | - |
| text-embedding-004 | vertexai | - | - | - |
| text-embedding-005 | vertexai | - | - | - |
| text-multilingual-embedding-002 | vertexai | - | - | - |
-| text-unicorn | vertexai | - | - | - |
-| textembedding-gecko | vertexai | - | - | - |
+| grok-2-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-3 | xai | 131072 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-3-mini | xai | 131072 | - | In: $0.30, Out: $0.50, Cache: $0.08 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-non-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-code-fast-1 | xai | 131072 | - | In: $2.00, Out: $10.00 |
-### Batch Processing (95)
+### Batch Processing (153)
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
+| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00 |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
+| claude-3-7-sonnet-20250219 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| embedding-001 | gemini | 2048 | 1 | - |
| embedding-gecko-001 | gemini | 1024 | 1 | - |
+| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
-| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-preview-image-generation | gemini | 32000 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
@@ -1999,18 +2010,20 @@ redirect_from:
| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-image-preview | gemini | 32768 | 32768 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
-| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.5-flash-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash-preview-tts | gemini | 8000 | 16000 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-preview-06-17 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-2.5-flash-lite-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro | gemini | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| gemini-2.5-pro-preview-tts | gemini | 8000 | 16000 | In: $1.25, Out: $10.00, Cache: $0.31 |
-| gemini-embedding-001 | gemini | 2048 | 1 | - |
-| gemini-embedding-exp | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
+| gemini-flash-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-lite-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
| text-embedding-004 | gemini | 2048 | 1 | - |
| codestral-2411-rc5 | mistral | 32768 | 8192 | - |
@@ -2045,8 +2058,8 @@ redirect_from:
| mistral-medium-latest | mistral | 32768 | 8192 | - |
| mistral-saba-2502 | mistral | 32768 | 8192 | - |
| mistral-saba-latest | mistral | 32768 | 8192 | - |
-| mistral-small | mistral | 32768 | 8192 | - |
| mistral-small-2312 | mistral | 32768 | 8192 | - |
+| mistral-small | mistral | 32768 | 8192 | - |
| mistral-small-2409 | mistral | 32768 | 8192 | - |
| mistral-small-2501 | mistral | 32768 | 8192 | - |
| mistral-small-2503 | mistral | 32768 | 8192 | - |
@@ -2067,9 +2080,55 @@ redirect_from:
| pixtral-12b-latest | mistral | 32768 | 8192 | - |
| pixtral-large-2411 | mistral | 32768 | 8192 | - |
| pixtral-large-latest | mistral | 32768 | 8192 | - |
-| text-embedding-3-large | openai | - | - | In: $0.13, Out: $0.13 |
-| text-embedding-3-small | openai | - | - | In: $0.02, Out: $0.02 |
-| text-embedding-ada-002 | openai | - | - | In: $0.10, Out: $0.10 |
+| gpt-3.5-turbo | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
+| gpt-4-0613 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
+| gpt-4 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
+| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
+| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
+| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| computer-use-preview-2025-03-11 | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| gpt-oss-120b | openai | 131072 | 131072 | - |
+| gpt-oss-20b | openai | 131072 | 131072 | - |
+| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-deep-research | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-pro | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o3-pro-2025-06-10 | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o4-mini-2025-04-16 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini-deep-research | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o4-mini-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| text-embedding-3-large | openai | - | - | In: $0.13 |
+| text-embedding-3-small | openai | - | - | In: $0.02 |
+| text-embedding-ada-002 | openai | - | - | In: $0.10 |
+| gemini-1.5-flash-002 | vertexai | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash | vertexai | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash-8b | vertexai | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-2.0-flash | vertexai | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-001 | vertexai | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-exp | vertexai | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
@@ -2081,53 +2140,53 @@ redirect_from:
## Models by Modality
-### Vision Models (224)
+### Vision Models (250)
Models that can process images:
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25, Cache: $0.30 |
-| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00, Cache: $1.00 |
+| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00 |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| claude-opus-4-20250514 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
-| claude-opus-4-1-20250805 | anthropic | 200000 | 32000 | In: $15.00, Out: $75.00, Cache: $18.75 |
+| claude-opus-4-20250514 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| claude-opus-4-1-20250805 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
-| claude-3-7-sonnet-20250219 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
-| claude-sonnet-4-20250514 | anthropic | 200000 | 64000 | In: $3.00, Out: $15.00, Cache: $3.75 |
+| claude-3-7-sonnet-20250219 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
+| claude-sonnet-4-20250514 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
-| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
-| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| us.anthropic.claude-opus-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-opus-4-1-20250805-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
| us.anthropic.claude-sonnet-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
+| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
@@ -2139,31 +2198,33 @@ Models that can process images:
| gemini-2.0-flash-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-exp | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-preview-image-generation | gemini | 32000 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
-| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-001 | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gemini-2.0-flash-lite | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| gemini-2.0-flash-lite-preview | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-flash-lite-preview-02-05 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.0-pro-exp-02-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
-| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-image-preview | gemini | 32768 | 32768 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.5-flash-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-preview-06-17 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-2.5-flash-lite-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro | gemini | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| gemini-2.5-pro-preview-06-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-03-25 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-05-06 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
-| gemini-embedding-exp | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-exp-1206 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
-| imagen-3.0-generate-002 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-preview-06-06 | gemini | 480 | 8192 | - |
-| imagen-4.0-ultra-generate-preview-06-06 | gemini | 480 | 8192 | - |
+| gemini-flash-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-lite-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-pro-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-robotics-er-1.5-preview | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
| mistral-large-pixtral-2411 | mistral | 32768 | 8192 | - |
| pixtral-12b | mistral | 32768 | 8192 | - |
@@ -2172,36 +2233,51 @@ Models that can process images:
| pixtral-large-2411 | mistral | 32768 | 8192 | - |
| pixtral-large-latest | mistral | 32768 | 8192 | - |
| chatgpt-4o-latest | openai | 128000 | 16384 | In: $5.00, Out: $15.00 |
-| gpt-4 | openai | 8192 | 8192 | In: $10.00, Out: $30.00 |
+| gpt-image-1 | openai | - | - | In: $5.00, Out: $40.00, Cache: $1.25 |
| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4-turbo-preview | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-5 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-chat-latest | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
-| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-chat-latest | openai | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-codex | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| omni-moderation-2024-09-26 | openai | - | - | - |
+| codex-mini-latest | openai | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
+| computer-use-preview-2025-03-11 | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| gpt-realtime-2025-08-28 | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
+| gpt-realtime | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
+| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-deep-research | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3-pro | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o3-pro-2025-06-10 | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o4-mini-2025-04-16 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini-deep-research | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o4-mini-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
| omni-moderation-latest | openai | - | - | - |
| allenai/molmo-7b-d | openrouter | 4096 | 4096 | In: $0.10, Out: $0.20 |
| amazon/nova-lite-v1 | openrouter | 300000 | 5120 | In: $0.06, Out: $0.24 |
@@ -2222,16 +2298,16 @@ Models that can process images:
| baidu/ernie-4.5-vl-424b-a47b | openrouter | 123000 | 16000 | In: $0.42, Out: $1.25 |
| bytedance/ui-tars-1.5-7b | openrouter | 128000 | 2048 | In: $0.10, Out: $0.20 |
| deepcogito/cogito-v2-preview-llama-109b-moe | openrouter | 32767 | - | In: $0.18, Out: $0.59 |
-| google/gemini-flash-1.5 | openrouter | 1000000 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
| google/gemini-flash-1.5-8b | openrouter | 1000000 | 8192 | In: $0.04, Out: $0.15, Cache: $0.01 |
-| google/gemini-pro-1.5 | openrouter | 2000000 | 8192 | In: $1.25, Out: $5.00 |
| google/gemini-2.0-flash-001 | openrouter | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.0-flash-exp:free | openrouter | 1048576 | 8192 | - |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| google/gemini-2.5-flash-image-preview | openrouter | 32768 | 8192 | In: $0.30, Out: $2.50 |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| google/gemini-2.5-flash-lite-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| google/gemini-2.5-flash-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview-05-06 | openrouter | 1048576 | 65535 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -2273,8 +2349,9 @@ Models that can process images:
| openai/gpt-4o:extended | openrouter | 128000 | 64000 | In: $6.00, Out: $18.00 |
| openai/gpt-4o-mini | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
| openai/gpt-4o-mini-2024-07-18 | openrouter | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
-| openai/gpt-5 | openrouter | 400000 | 128000 | In: $0.62, Out: $5.00, Cache: $0.06 |
+| openai/gpt-5 | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| openai/gpt-5-chat | openrouter | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| openai/gpt-5-codex | openrouter | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
| openai/gpt-5-mini | openrouter | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
| openai/gpt-5-nano | openrouter | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
| openai/o1 | openrouter | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
@@ -2292,12 +2369,15 @@ Models that can process images:
| qwen/qwen2.5-vl-32b-instruct | openrouter | 16384 | - | In: $0.04, Out: $0.14 |
| qwen/qwen2.5-vl-32b-instruct:free | openrouter | 8192 | - | - |
| qwen/qwen2.5-vl-72b-instruct | openrouter | 32768 | - | In: $0.07, Out: $0.28 |
-| qwen/qwen2.5-vl-72b-instruct:free | openrouter | 32768 | - | - |
+| qwen/qwen2.5-vl-72b-instruct:free | openrouter | 131072 | 2048 | - |
| qwen/qwen-2.5-vl-7b-instruct | openrouter | 32768 | - | In: $0.20, Out: $0.20 |
+| qwen/qwen3-vl-235b-a22b-instruct | openrouter | 131072 | 32768 | In: $0.30, Out: $1.50 |
+| qwen/qwen3-vl-235b-a22b-thinking | openrouter | 131072 | 32768 | In: $0.30, Out: $3.00 |
| stepfun-ai/step3 | openrouter | 65536 | 65536 | In: $0.57, Out: $1.42 |
| thudm/glm-4.1v-9b-thinking | openrouter | 65536 | 8000 | In: $0.04, Out: $0.14 |
| z-ai/glm-4.5v | openrouter | 65536 | 65536 | In: $0.50, Out: $1.80 |
| x-ai/grok-4 | openrouter | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| x-ai/grok-4-fast | openrouter | 2000000 | 30000 | In: $0.20, Out: $0.50, Cache: $0.05 |
| x-ai/grok-4-fast:free | openrouter | 2000000 | 30000 | - |
| gemini-1.5-flash | vertexai | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | vertexai | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
@@ -2311,23 +2391,28 @@ Models that can process images:
| gemini-2.5-flash | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite | vertexai | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-pro | vertexai | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
+| grok-2-image-1212 | xai | 131072 | - | In: $2.00, Out: $10.00 |
+| grok-2-vision-1212 | xai | 32768 | - | In: $2.00, Out: $10.00 |
+| grok-4-0709 | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-non-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
+| grok-4-fast-reasoning | xai | 256000 | - | In: $3.00, Out: $15.00, Cache: $0.75 |
-### Audio Input Models (63)
+### Audio Input Models (65)
Models that can process audio:
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
-| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
+| gemini-1.5-flash | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-002 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-001 | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
| gemini-1.5-flash-8b-latest | gemini | 1048576 | 8192 | In: $0.08, Out: $0.30, Cache: $0.02 |
-| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-001 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
+| gemini-1.5-pro | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-002 | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-1.5-pro-latest | gemini | 2097152 | 8192 | In: $1.25, Out: $5.00, Cache: $0.31 |
| gemini-2.0-flash | gemini | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
@@ -2341,31 +2426,33 @@ Models that can process audio:
| gemini-2.5-flash-preview-05-20 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-live-2.5-flash-preview | gemini | 1048576 | 8192 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-exp-native-audio-thinking-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
-| gemini-2.5-flash-preview-native-audio-dialog | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
+| gemini-2.5-flash-native-audio-preview-09-2025 | gemini | 128000 | 8000 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-flash-lite-06-17 | gemini | 1048576 | 65536 | In: $0.30, Out: $2.50, Cache: $0.08 |
| gemini-2.5-pro | gemini | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| gpt-4o-audio-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-transcribe | openai | 16000 | 2000 | In: $2.50, Out: $10.00 |
+| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-realtime-preview | openai | 16000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 16000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-transcribe | openai | 16000 | 2000 | In: $1.25, Out: $5.00 |
| gpt-4o-audio-preview-2024-10-01 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-audio-preview-2025-06-03 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-transcribe | openai | 16000 | 2000 | In: $1.25, Out: $5.00 |
-| gpt-4o-mini-tts | openai | - | - | In: $0.60, Out: $12.00 |
-| gpt-4o-transcribe | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-audio | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| gpt-audio-2025-08-28 | openai | 4096 | 16384 | In: $0.50, Out: $1.50 |
-| tts-1 | openai | - | - | In: $15.00, Out: $15.00 |
| tts-1-1106 | openai | - | - | In: $15.00, Out: $15.00 |
-| tts-1-hd | openai | - | - | In: $30.00, Out: $30.00 |
| tts-1-hd-1106 | openai | - | - | In: $30.00, Out: $30.00 |
-| whisper-1 | openai | - | - | In: $0.01, Out: $0.01 |
+| whisper-1 | openai | - | - | In: $0.01 |
+| gpt-audio | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-audio-2025-08-28 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-realtime | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
+| gpt-realtime-2025-08-28 | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
| google/gemini-2.0-flash-001 | openrouter | 1048576 | 8192 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.0-flash-lite-001 | openrouter | 1048576 | 8192 | In: $0.08, Out: $0.30 |
| google/gemini-2.5-flash | openrouter | 1048576 | 65535 | In: $0.30, Out: $2.50, Cache: $0.08 |
| google/gemini-2.5-flash-lite | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
| google/gemini-2.5-flash-lite-preview-06-17 | openrouter | 1048576 | 65535 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| google/gemini-2.5-flash-lite-preview-09-2025 | openrouter | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| google/gemini-2.5-pro | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview-05-06 | openrouter | 1048576 | 65535 | In: $1.25, Out: $10.00, Cache: $0.31 |
| google/gemini-2.5-pro-preview | openrouter | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
@@ -2384,38 +2471,44 @@ Models that can process audio:
| gemini-2.5-pro | vertexai | 1048576 | 65536 | In: $1.25, Out: $10.00, Cache: $0.31 |
-### PDF Models (81)
+### PDF Models (60)
Models that can process PDF documents:
| Model | Provider | Context | Max Output | Standard Pricing (per 1M tokens) |
| :-- | :-- | --: | --: | :-- |
+| claude-3-haiku-20240307 | anthropic | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| claude-3-5-haiku-20241022 | anthropic | 200000 | 8192 | In: $0.80, Out: $4.00 |
| claude-3-opus-20240229 | anthropic | 200000 | 4096 | In: $15.00, Out: $75.00 |
+| claude-opus-4-20250514 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| claude-opus-4-1-20250805 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20241022 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
| claude-3-5-sonnet-20240620 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
+| claude-3-7-sonnet-20250219 | anthropic | 200000 | 8192 | In: $3.00, Out: $15.00 |
+| claude-sonnet-4-20250514 | anthropic | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-v2:0:100k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-v2:1:200k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
-| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-v2:1:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
+| anthropic.claude-v2:0:18k | bedrock | 200000 | 4096 | In: $8.00, Out: $24.00 |
| anthropic.claude-3-haiku-20240307-v1:0:200k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
| anthropic.claude-3-haiku-20240307-v1:0:48k | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-3-haiku-20240307-v1:0 | bedrock | 200000 | 4096 | In: $0.25, Out: $1.25 |
+| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0 | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:12k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
-| anthropic.claude-3-opus-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-opus-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $15.00, Out: $75.00 |
| anthropic.claude-3-sonnet-20240229-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-sonnet-20240229-v1:0:28k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-haiku-20241022-v1:0 | bedrock | 200000 | 4096 | In: $0.80, Out: $4.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20240620-v1:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20240620-v1:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
-| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-3-5-sonnet-20241022-v2:0:51k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:200k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
+| anthropic.claude-3-5-sonnet-20241022-v2:0:18k | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| us.anthropic.claude-3-7-sonnet-20250219-v1:0 | bedrock | 200000 | 4096 | In: $3.00, Out: $15.00 |
| anthropic.claude-instant-v1:2:100k | bedrock | 200000 | 4096 | In: $0.80, Out: $2.40 |
| us.anthropic.claude-opus-4-20250514-v1:0 | bedrock | 200000 | 4096 | In: $0.10, Out: $0.20 |
@@ -2428,52 +2521,25 @@ Models that can process PDF documents:
| gemini-2.0-flash-thinking-exp | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-01-21 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
| gemini-2.0-flash-thinking-exp-1219 | gemini | 1048576 | 65536 | In: $0.10, Out: $0.40 |
+| gemini-2.5-flash-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-flash-lite-preview-06-17 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-2.5-flash-lite-preview-09-2025 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-06-05 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-03-25 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| gemini-2.5-pro-preview-05-06 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
-| gemini-embedding-exp | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-exp-1206 | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
-| imagen-3.0-generate-002 | gemini | 480 | 8192 | - |
-| imagen-4.0-generate-preview-06-06 | gemini | 480 | 8192 | - |
-| imagen-4.0-ultra-generate-preview-06-06 | gemini | 480 | 8192 | - |
+| gemini-flash-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-flash-lite-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-pro-latest | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
+| gemini-robotics-er-1.5-preview | gemini | 1048576 | 65536 | In: $0.08, Out: $0.30 |
| learnlm-2.0-flash-experimental | gemini | 1048576 | 32768 | In: $0.08, Out: $0.30 |
-| chatgpt-4o-latest | openai | 128000 | 16384 | In: $5.00, Out: $15.00 |
-| gpt-4 | openai | 8192 | 8192 | In: $10.00, Out: $30.00 |
-| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4-turbo-preview | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
-| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
-| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
-| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
-| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40 |
-| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-05-13 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
| gpt-4o-2024-11-20 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
-| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
-| gpt-5 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-chat-latest | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-mini-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| gpt-5-nano-2025-08-07 | openai | 128000 | 400000 | In: $1.25, Out: $10.00, Cache: $0.12 |
-| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00 |
-| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
-| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
| omni-moderation-2024-09-26 | openai | - | - | - |
-| omni-moderation-latest | openai | - | - | - |
-### Embedding Models (13)
+### Embedding Models (90)
Models that generate embeddings:
@@ -2481,15 +2547,92 @@ Models that generate embeddings:
| :-- | :-- | --: | --: | :-- |
| embedding-001 | gemini | 2048 | 1 | - |
| embedding-gecko-001 | gemini | 1024 | 1 | - |
-| gemini-embedding-001 | gemini | 2048 | 1 | - |
-| gemini-embedding-exp | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| gemini-embedding-exp-03-07 | gemini | 8192 | 1 | In: $0.00, Out: $0.00 |
| text-embedding-004 | gemini | 2048 | 1 | - |
-| codestral-embed | mistral | 32768 | 8192 | - |
| codestral-embed-2505 | mistral | 32768 | 8192 | - |
+| codestral-embed | mistral | 32768 | 8192 | - |
| mistral-embed | mistral | 32768 | 8192 | - |
| mistral-embed-2312 | mistral | 32768 | 8192 | - |
-| text-embedding-3-large | openai | - | - | In: $0.13, Out: $0.13 |
-| text-embedding-3-small | openai | - | - | In: $0.02, Out: $0.02 |
-| text-embedding-ada-002 | openai | - | - | In: $0.10, Out: $0.10 |
+| chatgpt-4o-latest | openai | 128000 | 16384 | In: $5.00, Out: $15.00 |
+| dall-e-2 | openai | - | - | - |
+| dall-e-3 | openai | - | - | - |
+| gpt-image-1 | openai | - | - | In: $5.00, Out: $40.00, Cache: $1.25 |
+| gpt-3.5-turbo | openai | 16385 | 4096 | In: $0.50, Out: $1.50 |
+| gpt-4-0613 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
+| gpt-4 | openai | 8192 | 8192 | In: $30.00, Out: $60.00 |
+| gpt-4-turbo | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
+| gpt-4-turbo-2024-04-09 | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
+| gpt-4-turbo-preview | openai | 128000 | 4096 | In: $10.00, Out: $30.00 |
+| gpt-4.1 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| gpt-4.1-2025-04-14 | openai | 1047576 | 32768 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| gpt-4.1-mini-2025-04-14 | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1-mini | openai | 1047576 | 32768 | In: $0.40, Out: $1.60, Cache: $0.10 |
+| gpt-4.1-nano | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.1-nano-2025-04-14 | openai | 1047576 | 32768 | In: $0.10, Out: $0.40, Cache: $0.02 |
+| gpt-4.5-preview | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4.5-preview-2025-02-27 | openai | 128000 | 16384 | In: $75.00, Out: $150.00, Cache: $37.50 |
+| gpt-4o-2024-08-06 | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o | openai | 128000 | 16384 | In: $2.50, Out: $10.00, Cache: $1.25 |
+| gpt-4o-audio-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-search-preview | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-4o-transcribe | openai | 16000 | 2000 | In: $2.50, Out: $10.00 |
+| gpt-4o-mini | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini-2024-07-18 | openai | 128000 | 16384 | In: $0.15, Out: $0.60, Cache: $0.08 |
+| gpt-4o-mini-audio-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-audio-preview-2024-12-17 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-realtime-preview-2024-12-17 | openai | 16000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-realtime-preview | openai | 16000 | 4096 | In: $0.60, Out: $2.40, Cache: $0.30 |
+| gpt-4o-mini-search-preview | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-search-preview-2025-03-11 | openai | 128000 | 16384 | In: $0.15, Out: $0.60 |
+| gpt-4o-mini-tts | openai | 2000 | - | In: $0.60, Out: $12.00 |
+| gpt-4o-mini-transcribe | openai | 16000 | 2000 | In: $1.25, Out: $5.00 |
+| gpt-5-2025-08-07 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5 | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-chat-latest | openai | 128000 | 16384 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| gpt-5-mini | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-mini-2025-08-07 | openai | 400000 | 128000 | In: $0.25, Out: $2.00, Cache: $0.02 |
+| gpt-5-nano-2025-08-07 | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-nano | openai | 400000 | 128000 | In: $0.05, Out: $0.40, Cache: $0.01 |
+| gpt-5-codex | openai | 400000 | 128000 | In: $1.25, Out: $10.00, Cache: $0.12 |
+| tts-1 | openai | - | - | In: $15.00 |
+| tts-1-hd | openai | - | - | In: $30.00 |
+| whisper-1 | openai | - | - | In: $0.01 |
+| babbage-002 | openai | - | 16384 | In: $0.40, Out: $0.40 |
+| codex-mini-latest | openai | 200000 | 100000 | In: $1.50, Out: $6.00, Cache: $0.38 |
+| computer-use-preview-2025-03-11 | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| computer-use-preview | openai | 8192 | 1024 | In: $3.00, Out: $12.00 |
+| davinci-002 | openai | - | 16384 | In: $2.00, Out: $2.00 |
+| gpt-audio-2025-08-28 | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-audio | openai | 128000 | 16384 | In: $2.50, Out: $10.00 |
+| gpt-oss-120b | openai | 131072 | 131072 | - |
+| gpt-oss-20b | openai | 131072 | 131072 | - |
+| gpt-realtime-2025-08-28 | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
+| gpt-realtime | openai | 32000 | 4096 | In: $4.00, Out: $16.00, Cache: $0.40 |
+| o1 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-2024-12-17 | openai | 200000 | 100000 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-preview | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-preview-2024-09-12 | openai | 128000 | 32768 | In: $15.00, Out: $60.00, Cache: $7.50 |
+| o1-mini | openai | 128000 | 65536 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o1-mini-2024-09-12 | openai | 128000 | 65536 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o1-pro-2025-03-19 | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o1-pro | openai | 200000 | 100000 | In: $150.00, Out: $600.00 |
+| o3-2025-04-16 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o3-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3-deep-research | openai | 200000 | 100000 | In: $10.00, Out: $40.00, Cache: $2.50 |
+| o3-mini-2025-01-31 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.55 |
+| o3-pro-2025-06-10 | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o3-pro | openai | 200000 | 100000 | In: $20.00, Out: $80.00 |
+| o4-mini-2025-04-16 | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini | openai | 200000 | 100000 | In: $1.10, Out: $4.40, Cache: $0.28 |
+| o4-mini-deep-research-2025-06-26 | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| o4-mini-deep-research | openai | 200000 | 100000 | In: $2.00, Out: $8.00, Cache: $0.50 |
+| omni-moderation-latest | openai | - | - | - |
+| text-embedding-3-large | openai | - | - | In: $0.13 |
+| text-embedding-3-small | openai | - | - | In: $0.02 |
+| text-embedding-ada-002 | openai | - | - | In: $0.10 |
+| text-moderation-latest | openai | - | 32768 | - |
+| text-moderation-stable | openai | - | 32768 | - |
diff --git a/lib/ruby_llm/aliases.json b/lib/ruby_llm/aliases.json
index fb99e03c6..59febddee 100644
--- a/lib/ruby_llm/aliases.json
+++ b/lib/ruby_llm/aliases.json
@@ -106,6 +106,14 @@
"gemini": "gemini-2.5-flash-lite-preview-06-17",
"openrouter": "google/gemini-2.5-flash-lite-preview-06-17"
},
+ "gemini-2.5-flash-lite-preview-09-2025": {
+ "gemini": "gemini-2.5-flash-lite-preview-09-2025",
+ "openrouter": "google/gemini-2.5-flash-lite-preview-09-2025"
+ },
+ "gemini-2.5-flash-preview-09-2025": {
+ "gemini": "gemini-2.5-flash-preview-09-2025",
+ "openrouter": "google/gemini-2.5-flash-preview-09-2025"
+ },
"gemini-2.5-pro": {
"gemini": "gemini-2.5-pro",
"openrouter": "google/gemini-2.5-pro",
@@ -115,10 +123,6 @@
"gemini": "gemini-2.5-pro-preview-05-06",
"openrouter": "google/gemini-2.5-pro-preview-05-06"
},
- "gemini-embedding-001": {
- "gemini": "gemini-embedding-001",
- "vertexai": "gemini-embedding-001"
- },
"gemini-exp-1206": {
"gemini": "gemini-exp-1206",
"vertexai": "gemini-exp-1206"
@@ -219,6 +223,10 @@
"openai": "gpt-5",
"openrouter": "openai/gpt-5"
},
+ "gpt-5-codex": {
+ "openai": "gpt-5-codex",
+ "openrouter": "openai/gpt-5-codex"
+ },
"gpt-5-mini": {
"openai": "gpt-5-mini",
"openrouter": "openai/gpt-5-mini"
@@ -227,6 +235,95 @@
"openai": "gpt-5-nano",
"openrouter": "openai/gpt-5-nano"
},
+ "gpt-oss-120b": {
+ "openai": "gpt-oss-120b",
+ "openrouter": "openai/gpt-oss-120b"
+ },
+ "gpt-oss-20b": {
+ "openai": "gpt-oss-20b",
+ "openrouter": "openai/gpt-oss-20b"
+ },
+ "grok-2": {
+ "xai": "grok-2-1212"
+ },
+ "grok-2-image": {
+ "xai": "grok-2-image-1212"
+ },
+ "grok-2-image-latest": {
+ "xai": "grok-2-image-1212"
+ },
+ "grok-2-latest": {
+ "xai": "grok-2-1212"
+ },
+ "grok-2-vision": {
+ "xai": "grok-2-vision-1212"
+ },
+ "grok-2-vision-latest": {
+ "xai": "grok-2-vision-1212"
+ },
+ "grok-3": {
+ "xai": "grok-3",
+ "openrouter": "x-ai/grok-3"
+ },
+ "grok-3-beta": {
+ "xai": "grok-3"
+ },
+ "grok-3-fast": {
+ "xai": "grok-3"
+ },
+ "grok-3-fast-beta": {
+ "xai": "grok-3"
+ },
+ "grok-3-fast-latest": {
+ "xai": "grok-3"
+ },
+ "grok-3-latest": {
+ "xai": "grok-3"
+ },
+ "grok-3-mini": {
+ "xai": "grok-3-mini",
+ "openrouter": "x-ai/grok-3-mini"
+ },
+ "grok-3-mini-beta": {
+ "xai": "grok-3-mini"
+ },
+ "grok-3-mini-fast": {
+ "xai": "grok-3-mini"
+ },
+ "grok-3-mini-fast-beta": {
+ "xai": "grok-3-mini"
+ },
+ "grok-3-mini-fast-latest": {
+ "xai": "grok-3-mini"
+ },
+ "grok-3-mini-latest": {
+ "xai": "grok-3-mini"
+ },
+ "grok-4": {
+ "xai": "grok-4-0709"
+ },
+ "grok-4-fast": {
+ "xai": "grok-4-fast-reasoning"
+ },
+ "grok-4-fast-non-reasoning-latest": {
+ "xai": "grok-4-fast-non-reasoning"
+ },
+ "grok-4-fast-reasoning-latest": {
+ "xai": "grok-4-fast-reasoning"
+ },
+ "grok-4-latest": {
+ "xai": "grok-4-0709"
+ },
+ "grok-code-fast": {
+ "xai": "grok-code-fast-1"
+ },
+ "grok-code-fast-1": {
+ "xai": "grok-code-fast-1",
+ "openrouter": "x-ai/grok-code-fast-1"
+ },
+ "grok-code-fast-1-0825": {
+ "xai": "grok-code-fast-1"
+ },
"o1": {
"openai": "o1",
"openrouter": "openai/o1"
diff --git a/lib/ruby_llm/models.json b/lib/ruby_llm/models.json
index 7ab399bab..0a57b5fbc 100644
--- a/lib/ruby_llm/models.json
+++ b/lib/ruby_llm/models.json
@@ -4,28 +4,34 @@
"name": "Claude Haiku 3.5",
"provider": "anthropic",
"family": "claude-3-5-haiku",
- "created_at": null,
+ "created_at": "2024-10-22 00:00:00 UTC",
"context_window": 200000,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "function_calling"
+ "streaming",
+ "function_calling",
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.8,
- "cached_input_per_million": 1.0,
"output_per_million": 4.0
+ },
+ "batch": {
+ "input_per_million": 0.4,
+ "output_per_million": 2.0
}
}
},
@@ -112,28 +118,38 @@
"name": "Claude Sonnet 3.7",
"provider": "anthropic",
"family": "claude-3-7-sonnet",
- "created_at": null,
+ "created_at": "2025-02-24 00:00:00 UTC",
"context_window": 200000,
- "max_output_tokens": 64000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "function_calling"
+ "streaming",
+ "function_calling",
+ "batch",
+ "reasoning",
+ "citations"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 3.0,
- "cached_input_per_million": 3.75,
- "output_per_million": 15.0
+ "output_per_million": 15.0,
+ "reasoning_output_per_million": 37.5
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5,
+ "reasoning_output_per_million": 18.75
}
}
},
@@ -144,28 +160,34 @@
"name": "Claude Haiku 3",
"provider": "anthropic",
"family": "claude-3-haiku",
- "created_at": null,
+ "created_at": "2024-03-07 00:00:00 UTC",
"context_window": 200000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "function_calling"
+ "streaming",
+ "function_calling",
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.25,
- "cached_input_per_million": 0.3,
"output_per_million": 1.25
+ },
+ "batch": {
+ "input_per_million": 0.125,
+ "output_per_million": 0.625
}
}
},
@@ -213,29 +235,34 @@
"id": "claude-opus-4-1-20250805",
"name": "Claude Opus 4.1",
"provider": "anthropic",
- "family": "claude-opus-4-1",
- "created_at": null,
+ "family": "claude-2",
+ "created_at": "2025-08-05 00:00:00 UTC",
"context_window": 200000,
- "max_output_tokens": 32000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "function_calling"
+ "streaming",
+ "reasoning"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 18.75,
- "output_per_million": 75.0
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
@@ -245,29 +272,34 @@
"id": "claude-opus-4-20250514",
"name": "Claude Opus 4",
"provider": "anthropic",
- "family": "claude-opus-4",
- "created_at": null,
+ "family": "claude-2",
+ "created_at": "2025-05-22 00:00:00 UTC",
"context_window": 200000,
- "max_output_tokens": 32000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "function_calling"
+ "streaming",
+ "reasoning"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "cached_input_per_million": 18.75,
- "output_per_million": 75.0
+ "input_per_million": 3.0,
+ "output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
@@ -277,29 +309,34 @@
"id": "claude-sonnet-4-20250514",
"name": "Claude Sonnet 4",
"provider": "anthropic",
- "family": "claude-sonnet-4",
- "created_at": null,
+ "family": "claude-2",
+ "created_at": "2025-05-22 00:00:00 UTC",
"context_window": 200000,
- "max_output_tokens": 64000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "text",
"image",
- "text"
+ "pdf"
],
"output": [
"text"
]
},
"capabilities": [
- "function_calling"
+ "streaming",
+ "reasoning"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 3.0,
- "cached_input_per_million": 3.75,
"output_per_million": 15.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 7.5
}
}
},
@@ -1759,35 +1796,6 @@
"owned_by": "deepseek"
}
},
- {
- "id": "aqa",
- "name": "Model that performs Attributed Question Answering.",
- "provider": "gemini",
- "family": "aqa",
- "created_at": null,
- "context_window": 7168,
- "max_output_tokens": 1024,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "text"
- ],
- "output": [
- "text"
- ]
- },
- "capabilities": [
- "streaming"
- ],
- "pricing": {},
- "metadata": {
- "version": "001",
- "description": "Model trained to return answers to questions that are grounded in provided sources, along with estimating answerable probability.",
- "supported_generation_methods": [
- "generateAnswer"
- ]
- }
- },
{
"id": "embedding-001",
"name": "Embedding 001",
@@ -1883,6 +1891,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -1892,6 +1901,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -1924,6 +1937,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -1933,6 +1947,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -1958,6 +1976,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -1967,6 +1986,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -2000,6 +2023,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -2009,6 +2033,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -2042,6 +2070,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -2051,6 +2080,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -2084,6 +2117,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -2093,6 +2127,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -2126,6 +2164,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -2135,6 +2174,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -3011,7 +3054,7 @@
"id": "gemini-2.5-flash-exp-native-audio-thinking-dialog",
"name": "Gemini 2.5 Flash Native Audio",
"provider": "gemini",
- "family": "gemini-2.5-flash-preview-native-audio-dialog",
+ "family": "gemini-2.5-flash-native-audio-preview-09-2025",
"created_at": null,
"context_window": 128000,
"max_output_tokens": 8000,
@@ -3221,45 +3264,45 @@
}
},
{
- "id": "gemini-2.5-flash-preview-05-20",
- "name": "Gemini 2.5 Flash",
+ "id": "gemini-2.5-flash-lite-preview-09-2025",
+ "name": "Gemini 2.5 Flash-Lite Preview Sep 2025",
"provider": "gemini",
- "family": "gemini-2.5-flash",
+ "family": "other",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
+ "text",
"image",
- "text"
+ "pdf",
+ "video"
],
"output": [
"text"
]
},
"capabilities": [
- "batch",
- "function_calling",
- "structured_output"
+ "streaming",
+ "structured_output",
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
},
"batch": {
- "input_per_million": 0.15,
- "output_per_million": 1.25
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "2.5-preview-05-20",
- "description": "Preview release (April 17th, 2025) of Gemini 2.5 Flash",
+ "version": "2.5-preview-09-25",
+ "description": "Preview release (Septempber 25th, 2025) of Gemini 2.5 Flash-Lite",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3269,10 +3312,10 @@
}
},
{
- "id": "gemini-2.5-flash-preview-native-audio-dialog",
+ "id": "gemini-2.5-flash-native-audio-preview-09-2025",
"name": "Gemini 2.5 Flash Native Audio",
"provider": "gemini",
- "family": "gemini-2.5-flash-preview-native-audio-dialog",
+ "family": "gemini-2.5-flash-native-audio-preview-09-2025",
"created_at": null,
"context_window": 128000,
"max_output_tokens": 8000,
@@ -3302,52 +3345,10 @@
"metadata": {}
},
{
- "id": "gemini-2.5-flash-preview-tts",
- "name": "Gemini 2.5 Flash Preview TTS",
- "provider": "gemini",
- "family": "gemini-2.5-flash-preview-tts",
- "created_at": null,
- "context_window": 8000,
- "max_output_tokens": 16000,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "text"
- ],
- "output": [
- "audio"
- ]
- },
- "capabilities": [
- "batch"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
- },
- "batch": {
- "input_per_million": 0.15,
- "output_per_million": 1.25
- }
- }
- },
- "metadata": {
- "version": "gemini-2.5-flash-exp-tts-2025-05-19",
- "description": "Gemini 2.5 Flash Preview TTS",
- "supported_generation_methods": [
- "countTokens",
- "generateContent"
- ]
- }
- },
- {
- "id": "gemini-2.5-pro",
- "name": "Gemini 2.5 Pro",
+ "id": "gemini-2.5-flash-preview-05-20",
+ "name": "Gemini 2.5 Flash",
"provider": "gemini",
- "family": "gemini-2.5-pro",
+ "family": "gemini-2.5-flash",
"created_at": null,
"context_window": 1048576,
"max_output_tokens": 65536,
@@ -3370,19 +3371,19 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.31,
- "output_per_million": 10.0
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
},
"batch": {
- "input_per_million": 0.625,
- "output_per_million": 5.0
+ "input_per_million": 0.15,
+ "output_per_million": 1.25
}
}
},
"metadata": {
- "version": "2.5",
- "description": "Stable release (June 17th, 2025) of Gemini 2.5 Pro",
+ "version": "2.5-preview-05-20",
+ "description": "Preview release (April 17th, 2025) of Gemini 2.5 Flash",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3392,8 +3393,8 @@
}
},
{
- "id": "gemini-2.5-pro-preview-03-25",
- "name": "Gemini 2.5 Pro Preview 03-25",
+ "id": "gemini-2.5-flash-preview-09-2025",
+ "name": "Gemini 2.5 Flash Preview Sep 2025",
"provider": "gemini",
"family": "other",
"created_at": null,
@@ -3415,6 +3416,7 @@
"streaming",
"function_calling",
"structured_output",
+ "batch",
"caching"
],
"pricing": {
@@ -3430,8 +3432,8 @@
}
},
"metadata": {
- "version": "2.5-preview-03-25",
- "description": "Gemini 2.5 Pro Preview 03-25",
+ "version": "Gemini 2.5 Flash Preview 09-2025",
+ "description": "Gemini 2.5 Flash Preview Sep 2025",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3441,46 +3443,87 @@
}
},
{
- "id": "gemini-2.5-pro-preview-05-06",
- "name": "Gemini 2.5 Pro Preview 05-06",
+ "id": "gemini-2.5-flash-preview-tts",
+ "name": "Gemini 2.5 Flash Preview TTS",
"provider": "gemini",
- "family": "other",
+ "family": "gemini-2.5-flash-preview-tts",
"created_at": null,
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "context_window": 8000,
+ "max_output_tokens": 16000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf",
- "video"
+ "text"
+ ],
+ "output": [
+ "audio"
+ ]
+ },
+ "capabilities": [
+ "batch"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
+ },
+ "batch": {
+ "input_per_million": 0.15,
+ "output_per_million": 1.25
+ }
+ }
+ },
+ "metadata": {
+ "version": "gemini-2.5-flash-exp-tts-2025-05-19",
+ "description": "Gemini 2.5 Flash Preview TTS",
+ "supported_generation_methods": [
+ "countTokens",
+ "generateContent"
+ ]
+ }
+ },
+ {
+ "id": "gemini-2.5-pro",
+ "name": "Gemini 2.5 Pro",
+ "provider": "gemini",
+ "family": "gemini-2.5-pro",
+ "created_at": null,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "audio",
+ "image",
+ "text"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "caching"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.31,
+ "output_per_million": 10.0
},
"batch": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
+ "input_per_million": 0.625,
+ "output_per_million": 5.0
}
}
},
"metadata": {
- "version": "2.5-preview-05-06",
- "description": "Preview release (May 6th, 2025) of Gemini 2.5 Pro",
+ "version": "2.5",
+ "description": "Stable release (June 17th, 2025) of Gemini 2.5 Pro",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3490,8 +3533,8 @@
}
},
{
- "id": "gemini-2.5-pro-preview-06-05",
- "name": "Gemini 2.5 Pro Preview",
+ "id": "gemini-2.5-pro-preview-03-25",
+ "name": "Gemini 2.5 Pro Preview 03-25",
"provider": "gemini",
"family": "other",
"created_at": null,
@@ -3528,8 +3571,8 @@
}
},
"metadata": {
- "version": "2.5-preview-06-05",
- "description": "Preview release (June 5th, 2025) of Gemini 2.5 Pro",
+ "version": "2.5-preview-03-25",
+ "description": "Gemini 2.5 Pro Preview 03-25",
"supported_generation_methods": [
"generateContent",
"countTokens",
@@ -3539,140 +3582,142 @@
}
},
{
- "id": "gemini-2.5-pro-preview-tts",
- "name": "Gemini 2.5 Pro Preview TTS",
+ "id": "gemini-2.5-pro-preview-05-06",
+ "name": "Gemini 2.5 Pro Preview 05-06",
"provider": "gemini",
- "family": "gemini-2.5-pro-preview-tts",
+ "family": "other",
"created_at": null,
- "context_window": 8000,
- "max_output_tokens": 16000,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "pdf",
+ "video"
],
"output": [
- "audio"
+ "text"
]
},
"capabilities": [
- "batch"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "cached_input_per_million": 0.31,
- "output_per_million": 10.0
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
},
"batch": {
- "input_per_million": 0.625,
- "output_per_million": 5.0
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "gemini-2.5-pro-preview-tts-2025-05-19",
- "description": "Gemini 2.5 Pro Preview TTS",
+ "version": "2.5-preview-05-06",
+ "description": "Preview release (May 6th, 2025) of Gemini 2.5 Pro",
"supported_generation_methods": [
+ "generateContent",
"countTokens",
- "generateContent"
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-embedding-001",
- "name": "Gemini Embedding 001",
+ "id": "gemini-2.5-pro-preview-06-05",
+ "name": "Gemini 2.5 Pro Preview",
"provider": "gemini",
- "family": "embedding1",
+ "family": "other",
"created_at": null,
- "context_window": 2048,
- "max_output_tokens": 1,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text",
+ "image",
+ "pdf",
"video"
],
"output": [
- "text",
- "embeddings"
+ "text"
]
},
"capabilities": [
"streaming",
- "batch"
+ "function_calling",
+ "structured_output",
+ "caching"
],
"pricing": {
- "embeddings": {
+ "text_tokens": {
"standard": {
- "input_per_million": 0.002
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "version": "001",
- "description": "Obtain a distributed representation of a text.",
+ "version": "2.5-preview-06-05",
+ "description": "Preview release (June 5th, 2025) of Gemini 2.5 Pro",
"supported_generation_methods": [
- "embedContent",
- "countTextTokens",
+ "generateContent",
"countTokens",
- "asyncBatchEmbedContent"
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemini-embedding-exp",
- "name": "Gemini Embedding Experimental",
+ "id": "gemini-2.5-pro-preview-tts",
+ "name": "Gemini 2.5 Pro Preview TTS",
"provider": "gemini",
- "family": "gemini_embedding_exp",
+ "family": "gemini-2.5-pro-preview-tts",
"created_at": null,
- "context_window": 8192,
- "max_output_tokens": 1,
+ "context_window": 8000,
+ "max_output_tokens": 16000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf",
- "video"
+ "text"
],
"output": [
- "text",
- "embeddings"
+ "audio"
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "batch",
- "caching"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.002,
- "output_per_million": 0.004
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.31,
+ "output_per_million": 10.0
},
"batch": {
- "input_per_million": 0.001,
- "output_per_million": 0.002
- }
- },
- "embeddings": {
- "standard": {
- "input_per_million": 0.002
+ "input_per_million": 0.625,
+ "output_per_million": 5.0
}
}
},
"metadata": {
- "version": "exp-03-07",
- "description": "Obtain a distributed representation of a text.",
+ "version": "gemini-2.5-pro-preview-tts-2025-05-19",
+ "description": "Gemini 2.5 Pro Preview TTS",
"supported_generation_methods": [
- "embedContent",
- "countTextTokens",
- "countTokens"
+ "countTokens",
+ "generateContent"
]
}
},
@@ -3781,58 +3826,79 @@
}
},
{
- "id": "gemini-live-2.5-flash-preview",
- "name": "Gemini 2.5 Flash Live",
+ "id": "gemini-flash-latest",
+ "name": "Gemini Flash Latest",
"provider": "gemini",
- "family": "gemini-live-2.5-flash-preview",
+ "family": "other",
"created_at": null,
"context_window": 1048576,
- "max_output_tokens": 8192,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "audio",
- "text"
+ "text",
+ "image",
+ "pdf",
+ "video"
],
"output": [
- "audio",
"text"
]
},
"capabilities": [
+ "streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "batch",
+ "caching"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "cached_input_per_million": 0.075,
- "output_per_million": 2.5
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
- "metadata": {}
+ "metadata": {
+ "version": "Gemini Flash Latest",
+ "description": "Latest release of Gemini Flash",
+ "supported_generation_methods": [
+ "generateContent",
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
+ ]
+ }
},
{
- "id": "gemma-3-12b-it",
- "name": "Gemma 3 12B",
+ "id": "gemini-flash-lite-latest",
+ "name": "Gemini Flash-Lite Latest",
"provider": "gemini",
"family": "other",
"created_at": null,
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "pdf",
+ "video"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "structured_output",
+ "batch"
],
"pricing": {
"text_tokens": {
@@ -3847,33 +3913,75 @@
}
},
"metadata": {
- "version": "001",
- "description": null,
+ "version": "Gemini Flash-Lite Latest",
+ "description": "Latest release of Gemini Flash-Lite",
"supported_generation_methods": [
"generateContent",
- "countTokens"
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemma-3-1b-it",
- "name": "Gemma 3 1B",
+ "id": "gemini-live-2.5-flash-preview",
+ "name": "Gemini 2.5 Flash Live",
"provider": "gemini",
- "family": "other",
+ "family": "gemini-live-2.5-flash-preview",
"created_at": null,
- "context_window": 32768,
+ "context_window": 1048576,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
"text"
],
"output": [
+ "audio",
"text"
]
},
"capabilities": [
- "streaming"
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.3,
+ "cached_input_per_million": 0.075,
+ "output_per_million": 2.5
+ }
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "id": "gemini-pro-latest",
+ "name": "Gemini Pro Latest",
+ "provider": "gemini",
+ "family": "other",
+ "created_at": null,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image",
+ "pdf",
+ "video"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "caching"
],
"pricing": {
"text_tokens": {
@@ -3888,33 +3996,41 @@
}
},
"metadata": {
- "version": "001",
- "description": null,
+ "version": "Gemini Pro Latest",
+ "description": "Latest release of Gemini Pro",
"supported_generation_methods": [
"generateContent",
- "countTokens"
+ "countTokens",
+ "createCachedContent",
+ "batchGenerateContent"
]
}
},
{
- "id": "gemma-3-27b-it",
- "name": "Gemma 3 27B",
+ "id": "gemini-robotics-er-1.5-preview",
+ "name": "Gemini Robotics-ER 1.5 Preview",
"provider": "gemini",
"family": "other",
"created_at": null,
- "context_window": 131072,
- "max_output_tokens": 8192,
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "pdf",
+ "video"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output",
+ "caching"
],
"pricing": {
"text_tokens": {
@@ -3929,8 +4045,8 @@
}
},
"metadata": {
- "version": "001",
- "description": null,
+ "version": "1.5-preview",
+ "description": "Gemini Robotics-ER 1.5 Preview",
"supported_generation_methods": [
"generateContent",
"countTokens"
@@ -3938,8 +4054,8 @@
}
},
{
- "id": "gemma-3-4b-it",
- "name": "Gemma 3 4B",
+ "id": "gemma-3-12b-it",
+ "name": "Gemma 3 12B",
"provider": "gemini",
"family": "other",
"created_at": null,
@@ -3979,13 +4095,13 @@
}
},
{
- "id": "gemma-3n-e2b-it",
- "name": "Gemma 3n E2B",
+ "id": "gemma-3-1b-it",
+ "name": "Gemma 3 1B",
"provider": "gemini",
"family": "other",
"created_at": null,
- "context_window": 8192,
- "max_output_tokens": 2048,
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -4020,13 +4136,13 @@
}
},
{
- "id": "gemma-3n-e4b-it",
- "name": "Gemma 3n E4B",
+ "id": "gemma-3-27b-it",
+ "name": "Gemma 3 27B",
"provider": "gemini",
"family": "other",
"created_at": null,
- "context_window": 8192,
- "max_output_tokens": 2048,
+ "context_window": 131072,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -4061,95 +4177,125 @@
}
},
{
- "id": "imagen-3.0-generate-002",
- "name": "Imagen 3.0",
+ "id": "gemma-3-4b-it",
+ "name": "Gemma 3 4B",
"provider": "gemini",
- "family": "imagen3",
+ "family": "other",
"created_at": null,
- "context_window": 480,
+ "context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "image"
+ "text"
]
},
"capabilities": [
"streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
+ }
+ }
+ },
"metadata": {
- "version": "002",
- "description": "Vertex served Imagen 3.0 002 model",
+ "version": "001",
+ "description": null,
"supported_generation_methods": [
- "predict"
+ "generateContent",
+ "countTokens"
]
}
},
{
- "id": "imagen-4.0-generate-preview-06-06",
- "name": "Imagen 4 (Preview)",
+ "id": "gemma-3n-e2b-it",
+ "name": "Gemma 3n E2B",
"provider": "gemini",
"family": "other",
"created_at": null,
- "context_window": 480,
- "max_output_tokens": 8192,
+ "context_window": 8192,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "image"
+ "text"
]
},
"capabilities": [
"streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
+ }
+ }
+ },
"metadata": {
- "version": "01",
- "description": "Vertex served Imagen 4.0 model",
+ "version": "001",
+ "description": null,
"supported_generation_methods": [
- "predict"
+ "generateContent",
+ "countTokens"
]
}
},
{
- "id": "imagen-4.0-ultra-generate-preview-06-06",
- "name": "Imagen 4 Ultra (Preview)",
+ "id": "gemma-3n-e4b-it",
+ "name": "Gemma 3n E4B",
"provider": "gemini",
"family": "other",
"created_at": null,
- "context_window": 480,
- "max_output_tokens": 8192,
+ "context_window": 8192,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
- "image"
+ "text"
]
},
"capabilities": [
"streaming"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
+ }
+ }
+ },
"metadata": {
- "version": "01",
- "description": "Vertex served Imagen 4.0 ultra model",
+ "version": "001",
+ "description": null,
"supported_generation_methods": [
- "predict"
+ "generateContent",
+ "countTokens"
]
}
},
@@ -6216,11 +6362,11 @@
},
{
"id": "babbage-002",
- "name": "Babbage 002",
+ "name": "babbage-002",
"provider": "openai",
- "family": "babbage",
- "created_at": "2023-08-21 18:16:55 +0200",
- "context_window": 4096,
+ "family": "babbage-002",
+ "created_at": null,
+ "context_window": null,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
@@ -6228,12 +6374,11 @@
"text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
@@ -6249,27 +6394,24 @@
},
{
"id": "chatgpt-4o-latest",
- "name": "ChatGPT-4o Latest",
+ "name": "ChatGPT-4o",
"provider": "openai",
- "family": "chatgpt4o",
- "created_at": "2024-08-13 04:12:11 +0200",
+ "family": "chatgpt-4o-latest",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming",
- "structured_output"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
@@ -6285,30 +6427,33 @@
},
{
"id": "codex-mini-latest",
- "name": "Codex Mini Latest",
+ "name": "codex-mini-latest",
"provider": "openai",
- "family": "other",
- "created_at": "2025-05-08 05:00:57 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "codex-mini-latest",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 1.5,
+ "cached_input_per_million": 0.375,
+ "output_per_million": 6.0
}
}
},
@@ -6318,31 +6463,37 @@
}
},
{
- "id": "dall-e-2",
- "name": "DALL-E-2",
+ "id": "computer-use-preview",
+ "name": "computer-use-preview",
"provider": "openai",
- "family": "dall_e",
- "created_at": "2023-11-01 01:22:57 +0100",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "computer-use-preview",
+ "created_at": null,
+ "context_window": 8192,
+ "max_output_tokens": 1024,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
- "text",
- "image"
+ "embeddings",
+ "text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 3.0,
+ "output_per_million": 12.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 6.0
}
}
},
@@ -6352,31 +6503,37 @@
}
},
{
- "id": "dall-e-3",
- "name": "DALL-E-3",
+ "id": "computer-use-preview-2025-03-11",
+ "name": "computer-use-preview",
"provider": "openai",
- "family": "dall_e",
- "created_at": "2023-10-31 21:46:29 +0100",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "computer-use-preview",
+ "created_at": null,
+ "context_window": 8192,
+ "max_output_tokens": 1024,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
- "text",
- "image"
+ "embeddings",
+ "text"
]
},
"capabilities": [
- "streaming"
+ "batch",
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 3.0,
+ "output_per_million": 12.0
+ },
+ "batch": {
+ "input_per_million": 1.5,
+ "output_per_million": 6.0
}
}
},
@@ -6386,29 +6543,78 @@
}
},
{
- "id": "davinci-002",
- "name": "Davinci 002",
+ "id": "dall-e-2",
+ "name": "DALL·E 2",
"provider": "openai",
- "family": "davinci",
- "created_at": "2023-08-21 18:11:41 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "dall-e-2",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
- "text"
+ "embeddings",
+ "image"
]
},
- "capabilities": [
- "streaming"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 2.0,
+ "capabilities": [],
+ "pricing": {},
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
+ },
+ {
+ "id": "dall-e-3",
+ "name": "DALL·E 3",
+ "provider": "openai",
+ "family": "dall-e-3",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "image"
+ ]
+ },
+ "capabilities": [],
+ "pricing": {},
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
+ },
+ {
+ "id": "davinci-002",
+ "name": "davinci-002",
+ "provider": "openai",
+ "family": "davinci-002",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": 16384,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "text"
+ ]
+ },
+ "capabilities": [],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
"output_per_million": 2.0
}
}
@@ -6422,8 +6628,8 @@
"id": "gpt-3.5-turbo",
"name": "GPT-3.5 Turbo",
"provider": "openai",
- "family": "gpt35_turbo",
- "created_at": "2023-02-28 19:56:42 +0100",
+ "family": "gpt-3.5-turbo",
+ "created_at": null,
"context_window": 16385,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
@@ -6432,17 +6638,22 @@
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.5,
"output_per_million": 1.5
+ },
+ "batch": {
+ "input_per_million": 0.25,
+ "output_per_million": 0.75
}
}
},
@@ -6620,29 +6831,31 @@
"id": "gpt-4",
"name": "GPT-4",
"provider": "openai",
- "family": "gpt4",
- "created_at": "2023-06-27 18:13:31 +0200",
+ "family": "gpt-4",
+ "created_at": null,
"context_window": 8192,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "function_calling"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 10.0,
+ "input_per_million": 30.0,
+ "output_per_million": 60.0
+ },
+ "batch": {
+ "input_per_million": 15.0,
"output_per_million": 30.0
}
}
@@ -6687,29 +6900,34 @@
},
{
"id": "gpt-4-0613",
- "name": "GPT-4 0613",
+ "name": "GPT-4",
"provider": "openai",
- "family": "other",
- "created_at": "2023-06-12 18:54:56 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "gpt-4",
+ "created_at": null,
+ "context_window": 8192,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 30.0,
+ "output_per_million": 60.0
+ },
+ "batch": {
+ "input_per_million": 15.0,
+ "output_per_million": 30.0
}
}
},
@@ -6755,23 +6973,23 @@
"id": "gpt-4-turbo",
"name": "GPT-4 Turbo",
"provider": "openai",
- "family": "gpt4_turbo",
- "created_at": "2024-04-06 01:57:21 +0200",
+ "family": "gpt-4-turbo",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling"
],
"pricing": {
@@ -6779,6 +6997,10 @@
"standard": {
"input_per_million": 10.0,
"output_per_million": 30.0
+ },
+ "batch": {
+ "input_per_million": 5.0,
+ "output_per_million": 15.0
}
}
},
@@ -6789,25 +7011,25 @@
},
{
"id": "gpt-4-turbo-2024-04-09",
- "name": "GPT-4 Turbo 20240409",
+ "name": "GPT-4 Turbo",
"provider": "openai",
- "family": "gpt4_turbo",
- "created_at": "2024-04-08 20:41:17 +0200",
+ "family": "gpt-4-turbo",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling"
],
"pricing": {
@@ -6815,6 +7037,10 @@
"standard": {
"input_per_million": 10.0,
"output_per_million": 30.0
+ },
+ "batch": {
+ "input_per_million": 5.0,
+ "output_per_million": 15.0
}
}
},
@@ -6827,25 +7053,21 @@
"id": "gpt-4-turbo-preview",
"name": "GPT-4 Turbo Preview",
"provider": "openai",
- "family": "gpt4_turbo",
- "created_at": "2024-01-23 20:22:57 +0100",
+ "family": "gpt-4-turbo-preview",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming",
- "function_calling"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
@@ -6863,23 +7085,23 @@
"id": "gpt-4.1",
"name": "GPT-4.1",
"provider": "openai",
- "family": "gpt41",
- "created_at": "2025-04-10 22:22:22 +0200",
+ "family": "gpt-4.1",
+ "created_at": null,
"context_window": 1047576,
"max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
@@ -6887,8 +7109,12 @@
"text_tokens": {
"standard": {
"input_per_million": 2.0,
- "output_per_million": 8.0,
- "cached_input_per_million": 0.5
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
+ },
+ "batch": {
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
@@ -6899,25 +7125,25 @@
},
{
"id": "gpt-4.1-2025-04-14",
- "name": "GPT-4.1 20250414",
+ "name": "GPT-4.1",
"provider": "openai",
- "family": "gpt41",
- "created_at": "2025-04-10 22:09:06 +0200",
+ "family": "gpt-4.1",
+ "created_at": null,
"context_window": 1047576,
"max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
@@ -6925,8 +7151,12 @@
"text_tokens": {
"standard": {
"input_per_million": 2.0,
- "output_per_million": 8.0,
- "cached_input_per_million": 0.5
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
+ },
+ "batch": {
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
@@ -6937,25 +7167,25 @@
},
{
"id": "gpt-4.1-mini",
- "name": "GPT-4.1 Mini",
+ "name": "GPT-4.1 mini",
"provider": "openai",
- "family": "gpt41_mini",
- "created_at": "2025-04-10 22:49:33 +0200",
+ "family": "gpt-4.1-mini",
+ "created_at": null,
"context_window": 1047576,
"max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
@@ -6963,8 +7193,12 @@
"text_tokens": {
"standard": {
"input_per_million": 0.4,
- "output_per_million": 1.6,
- "cached_input_per_million": 0.1
+ "cached_input_per_million": 0.1,
+ "output_per_million": 1.6
+ },
+ "batch": {
+ "input_per_million": 0.2,
+ "output_per_million": 0.8
}
}
},
@@ -6975,25 +7209,25 @@
},
{
"id": "gpt-4.1-mini-2025-04-14",
- "name": "GPT-4.1 Mini 20250414",
+ "name": "GPT-4.1 mini",
"provider": "openai",
- "family": "gpt41_mini",
- "created_at": "2025-04-10 22:39:07 +0200",
+ "family": "gpt-4.1-mini",
+ "created_at": null,
"context_window": 1047576,
"max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
@@ -7001,8 +7235,12 @@
"text_tokens": {
"standard": {
"input_per_million": 0.4,
- "output_per_million": 1.6,
- "cached_input_per_million": 0.1
+ "cached_input_per_million": 0.1,
+ "output_per_million": 1.6
+ },
+ "batch": {
+ "input_per_million": 0.2,
+ "output_per_million": 0.8
}
}
},
@@ -7013,25 +7251,25 @@
},
{
"id": "gpt-4.1-nano",
- "name": "GPT-4.1 Nano",
+ "name": "GPT-4.1 nano",
"provider": "openai",
- "family": "gpt41_nano",
- "created_at": "2025-04-10 23:48:27 +0200",
+ "family": "gpt-4.1-nano",
+ "created_at": null,
"context_window": 1047576,
"max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
@@ -7039,7 +7277,12 @@
"text_tokens": {
"standard": {
"input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
"output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
@@ -7050,25 +7293,25 @@
},
{
"id": "gpt-4.1-nano-2025-04-14",
- "name": "GPT-4.1 Nano 20250414",
+ "name": "GPT-4.1 nano",
"provider": "openai",
- "family": "gpt41_nano",
- "created_at": "2025-04-10 23:37:05 +0200",
+ "family": "gpt-4.1-nano",
+ "created_at": null,
"context_window": 1047576,
"max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
@@ -7076,7 +7319,12 @@
"text_tokens": {
"standard": {
"input_per_million": 0.1,
+ "cached_input_per_million": 0.025,
"output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.05,
+ "output_per_million": 0.2
}
}
},
@@ -7086,100 +7334,104 @@
}
},
{
- "id": "gpt-4o",
- "name": "GPT-4o",
+ "id": "gpt-4.5-preview",
+ "name": "GPT-4.5 Preview (Deprecated)",
"provider": "openai",
- "family": "gpt4o",
- "created_at": "2024-05-10 20:50:49 +0200",
+ "family": "gpt-4.5-preview",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 75.0,
+ "cached_input_per_million": 37.5,
+ "output_per_million": 150.0
+ },
+ "batch": {
+ "input_per_million": 37.5,
+ "output_per_million": 75.0
}
}
},
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
+ "metadata": {}
},
{
- "id": "gpt-4o-2024-05-13",
- "name": "GPT-4o 20240513",
+ "id": "gpt-4.5-preview-2025-02-27",
+ "name": "GPT-4.5 Preview (Deprecated)",
"provider": "openai",
- "family": "gpt4o",
- "created_at": "2024-05-10 21:08:52 +0200",
+ "family": "gpt-4.5-preview",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 75.0,
+ "cached_input_per_million": 37.5,
+ "output_per_million": 150.0
+ },
+ "batch": {
+ "input_per_million": 37.5,
+ "output_per_million": 75.0
}
}
},
- "metadata": {
- "object": "model",
- "owned_by": "system"
- }
+ "metadata": {}
},
{
- "id": "gpt-4o-2024-08-06",
- "name": "GPT-4o 20240806",
+ "id": "gpt-4o",
+ "name": "GPT-4o",
"provider": "openai",
- "family": "gpt4o",
- "created_at": "2024-08-05 01:38:39 +0200",
+ "family": "gpt-4o",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
@@ -7187,7 +7439,12 @@
"text_tokens": {
"standard": {
"input_per_million": 2.5,
+ "cached_input_per_million": 1.25,
"output_per_million": 10.0
+ },
+ "batch": {
+ "input_per_million": 1.25,
+ "output_per_million": 5.0
}
}
},
@@ -7197,11 +7454,11 @@
}
},
{
- "id": "gpt-4o-2024-11-20",
- "name": "GPT-4o 20241120",
+ "id": "gpt-4o-2024-05-13",
+ "name": "GPT-4o 20240513",
"provider": "openai",
"family": "gpt4o",
- "created_at": "2025-02-12 04:39:03 +0100",
+ "created_at": "2024-05-10 21:08:52 +0200",
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
@@ -7234,28 +7491,106 @@
}
},
{
- "id": "gpt-4o-audio-preview",
- "name": "GPT-4o-Audio Preview",
+ "id": "gpt-4o-2024-08-06",
+ "name": "GPT-4o",
"provider": "openai",
- "family": "gpt4o_audio",
- "created_at": "2024-09-27 20:07:23 +0200",
+ "family": "gpt-4o",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "image",
+ "text"
],
"output": [
- "text",
- "audio"
+ "embeddings",
+ "text"
]
},
"capabilities": [
- "streaming",
- "speech_generation",
- "transcription"
+ "batch",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.5,
+ "cached_input_per_million": 1.25,
+ "output_per_million": 10.0
+ },
+ "batch": {
+ "input_per_million": 1.25,
+ "output_per_million": 5.0
+ }
+ }
+ },
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
+ },
+ {
+ "id": "gpt-4o-2024-11-20",
+ "name": "GPT-4o 20241120",
+ "provider": "openai",
+ "family": "gpt4o",
+ "created_at": "2025-02-12 04:39:03 +0100",
+ "context_window": 128000,
+ "max_output_tokens": 16384,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image",
+ "pdf"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
+ }
+ }
+ },
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
+ },
+ {
+ "id": "gpt-4o-audio-preview",
+ "name": "GPT-4o Audio",
+ "provider": "openai",
+ "family": "gpt-4o-audio-preview",
+ "created_at": null,
+ "context_window": 128000,
+ "max_output_tokens": 16384,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "audio",
+ "text"
+ ],
+ "output": [
+ "audio",
+ "embeddings",
+ "text"
+ ]
+ },
+ "capabilities": [
+ "function_calling"
],
"pricing": {
"text_tokens": {
@@ -7383,25 +7718,25 @@
},
{
"id": "gpt-4o-mini",
- "name": "GPT-4o-Mini",
+ "name": "GPT-4o mini",
"provider": "openai",
- "family": "gpt4o_mini",
- "created_at": "2024-07-17 01:32:21 +0200",
+ "family": "gpt-4o-mini",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
@@ -7409,7 +7744,12 @@
"text_tokens": {
"standard": {
"input_per_million": 0.15,
+ "cached_input_per_million": 0.075,
"output_per_million": 0.6
+ },
+ "batch": {
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
}
}
},
@@ -7420,25 +7760,25 @@
},
{
"id": "gpt-4o-mini-2024-07-18",
- "name": "GPT-4o-Mini 20240718",
+ "name": "GPT-4o mini",
"provider": "openai",
- "family": "gpt4o_mini",
- "created_at": "2024-07-17 01:31:57 +0200",
+ "family": "gpt-4o-mini",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
"structured_output"
],
@@ -7446,7 +7786,12 @@
"text_tokens": {
"standard": {
"input_per_million": 0.15,
+ "cached_input_per_million": 0.075,
"output_per_million": 0.6
+ },
+ "batch": {
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
}
}
},
@@ -7457,27 +7802,26 @@
},
{
"id": "gpt-4o-mini-audio-preview",
- "name": "GPT-4o-Mini Audio Preview",
+ "name": "GPT-4o mini Audio",
"provider": "openai",
- "family": "gpt4o_mini_audio",
- "created_at": "2024-12-16 23:17:04 +0100",
+ "family": "gpt-4o-mini-audio-preview",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "audio",
+ "text"
],
"output": [
- "text",
- "audio"
+ "audio",
+ "embeddings",
+ "text"
]
},
"capabilities": [
- "streaming",
- "speech_generation",
- "transcription"
+ "function_calling"
],
"pricing": {
"text_tokens": {
@@ -7494,27 +7838,26 @@
},
{
"id": "gpt-4o-mini-audio-preview-2024-12-17",
- "name": "GPT-4o-Mini Audio Preview 20241217",
+ "name": "GPT-4o mini Audio",
"provider": "openai",
- "family": "gpt4o_mini_audio",
- "created_at": "2024-12-13 19:52:00 +0100",
+ "family": "gpt-4o-mini-audio-preview",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "audio",
+ "text"
],
"output": [
- "text",
- "audio"
+ "audio",
+ "embeddings",
+ "text"
]
},
"capabilities": [
- "streaming",
- "speech_generation",
- "transcription"
+ "function_calling"
],
"pricing": {
"text_tokens": {
@@ -7531,28 +7874,32 @@
},
{
"id": "gpt-4o-mini-realtime-preview",
- "name": "GPT-4o-Mini Realtime Preview",
+ "name": "GPT-4o mini Realtime",
"provider": "openai",
- "family": "gpt4o_mini_realtime",
- "created_at": "2024-12-16 23:16:20 +0100",
- "context_window": 128000,
+ "family": "gpt-4o-mini-realtime-preview",
+ "created_at": null,
+ "context_window": 16000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
"text"
],
"output": [
+ "audio",
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.6,
+ "cached_input_per_million": 0.3,
"output_per_million": 2.4
}
}
@@ -7564,28 +7911,32 @@
},
{
"id": "gpt-4o-mini-realtime-preview-2024-12-17",
- "name": "GPT-4o-Mini Realtime Preview 20241217",
+ "name": "GPT-4o mini Realtime",
"provider": "openai",
- "family": "gpt4o_mini_realtime",
- "created_at": "2024-12-13 18:56:41 +0100",
- "context_window": 128000,
+ "family": "gpt-4o-mini-realtime-preview",
+ "created_at": null,
+ "context_window": 16000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
"text"
],
"output": [
+ "audio",
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.6,
+ "cached_input_per_million": 0.3,
"output_per_million": 2.4
}
}
@@ -7597,11 +7948,11 @@
},
{
"id": "gpt-4o-mini-search-preview",
- "name": "GPT-4o-Mini Search Preview",
+ "name": "GPT-4o mini Search Preview",
"provider": "openai",
- "family": "other",
- "created_at": "2025-03-08 00:46:01 +0100",
- "context_window": 4096,
+ "family": "gpt-4o-mini-search-preview",
+ "created_at": null,
+ "context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
@@ -7609,17 +7960,18 @@
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
}
}
},
@@ -7630,11 +7982,11 @@
},
{
"id": "gpt-4o-mini-search-preview-2025-03-11",
- "name": "GPT-4o-Mini Search Preview 20250311",
+ "name": "GPT-4o mini Search Preview",
"provider": "openai",
- "family": "other",
- "created_at": "2025-03-08 00:40:58 +0100",
- "context_window": 4096,
+ "family": "gpt-4o-mini-search-preview",
+ "created_at": null,
+ "context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
@@ -7642,17 +7994,18 @@
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
}
}
},
@@ -7663,25 +8016,24 @@
},
{
"id": "gpt-4o-mini-transcribe",
- "name": "GPT-4o-Mini Transcribe",
+ "name": "GPT-4o mini Transcribe",
"provider": "openai",
- "family": "gpt4o_mini_transcribe",
- "created_at": "2025-03-15 20:56:36 +0100",
+ "family": "gpt-4o-mini-transcribe",
+ "created_at": null,
"context_window": 16000,
"max_output_tokens": 2000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "audio",
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
@@ -7697,26 +8049,23 @@
},
{
"id": "gpt-4o-mini-tts",
- "name": "GPT-4o-Mini Tts",
+ "name": "GPT-4o mini TTS",
"provider": "openai",
- "family": "gpt4o_mini_tts",
- "created_at": "2025-03-19 18:05:59 +0100",
- "context_window": null,
+ "family": "gpt-4o-mini-tts",
+ "created_at": null,
+ "context_window": 2000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "text"
],
"output": [
- "text",
- "audio"
+ "audio",
+ "embeddings"
]
},
- "capabilities": [
- "streaming"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
@@ -7866,23 +8215,22 @@
"id": "gpt-4o-search-preview",
"name": "GPT-4o Search Preview",
"provider": "openai",
- "family": "gpt4o_search",
- "created_at": "2025-03-08 00:05:20 +0100",
+ "family": "gpt-4o-search-preview",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "structured_output"
],
"pricing": {
"text_tokens": {
@@ -7899,25 +8247,24 @@
},
{
"id": "gpt-4o-search-preview-2025-03-11",
- "name": "GPT-4o Search Preview 20250311",
+ "name": "GPT-4o Search Preview",
"provider": "openai",
- "family": "gpt4o_search",
- "created_at": "2025-03-07 23:56:10 +0100",
+ "family": "gpt-4o-search-preview",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "structured_output"
],
"pricing": {
"text_tokens": {
@@ -7934,25 +8281,24 @@
},
{
"id": "gpt-4o-transcribe",
- "name": "GPT-4o-Transcribe",
+ "name": "GPT-4o Transcribe",
"provider": "openai",
- "family": "gpt4o_transcribe",
- "created_at": "2025-03-15 20:54:23 +0100",
- "context_window": 128000,
- "max_output_tokens": 16384,
+ "family": "gpt-4o-transcribe",
+ "created_at": null,
+ "context_window": 16000,
+ "max_output_tokens": 2000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "audio",
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
@@ -7970,33 +8316,36 @@
"id": "gpt-5",
"name": "GPT-5",
"provider": "openai",
- "family": "gpt5",
- "created_at": "2025-08-05 22:29:37 +0200",
- "context_window": 128000,
- "max_output_tokens": 400000,
+ "family": "gpt-5",
+ "created_at": null,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.125
+ "cached_input_per_million": 0.125,
+ "output_per_million": 10.0
+ },
+ "batch": {
+ "input_per_million": 0.625,
+ "output_per_million": 5.0
}
}
},
@@ -8007,35 +8356,38 @@
},
{
"id": "gpt-5-2025-08-07",
- "name": "GPT-5 20250807",
+ "name": "GPT-5",
"provider": "openai",
- "family": "gpt5",
- "created_at": "2025-08-01 21:09:20 +0200",
- "context_window": 128000,
- "max_output_tokens": 400000,
+ "family": "gpt-5",
+ "created_at": null,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.125
+ "cached_input_per_million": 0.125,
+ "output_per_million": 10.0
+ },
+ "batch": {
+ "input_per_million": 0.625,
+ "output_per_million": 5.0
}
}
},
@@ -8046,35 +8398,67 @@
},
{
"id": "gpt-5-chat-latest",
- "name": "GPT-5 Chat Latest",
+ "name": "GPT-5 Chat",
"provider": "openai",
- "family": "gpt5",
- "created_at": "2025-08-01 20:35:06 +0200",
+ "family": "gpt-5-chat-latest",
+ "created_at": null,
"context_window": 128000,
- "max_output_tokens": 400000,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "text"
+ ]
+ },
+ "capabilities": [],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.25,
+ "cached_input_per_million": 0.125,
+ "output_per_million": 10.0
+ }
+ }
+ },
+ "metadata": {
+ "object": "model",
+ "owned_by": "system"
+ }
+ },
+ {
+ "id": "gpt-5-codex",
+ "name": "GPT-5-Codex",
+ "provider": "openai",
+ "family": "gpt-5-codex",
+ "created_at": null,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "image",
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.125
+ "cached_input_per_million": 0.125,
+ "output_per_million": 10.0
}
}
},
@@ -8085,35 +8469,38 @@
},
{
"id": "gpt-5-mini",
- "name": "GPT-5 Mini",
+ "name": "GPT-5 mini",
"provider": "openai",
- "family": "gpt5",
- "created_at": "2025-08-05 22:32:08 +0200",
- "context_window": 128000,
- "max_output_tokens": 400000,
+ "family": "gpt-5-mini",
+ "created_at": null,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.125
+ "input_per_million": 0.25,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 2.0
+ },
+ "batch": {
+ "input_per_million": 0.125,
+ "output_per_million": 1.0
}
}
},
@@ -8124,35 +8511,38 @@
},
{
"id": "gpt-5-mini-2025-08-07",
- "name": "GPT-5 Mini 20250807",
+ "name": "GPT-5 mini",
"provider": "openai",
- "family": "gpt5",
- "created_at": "2025-08-05 22:31:07 +0200",
- "context_window": 128000,
- "max_output_tokens": 400000,
+ "family": "gpt-5-mini",
+ "created_at": null,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.125
+ "input_per_million": 0.25,
+ "cached_input_per_million": 0.025,
+ "output_per_million": 2.0
+ },
+ "batch": {
+ "input_per_million": 0.125,
+ "output_per_million": 1.0
}
}
},
@@ -8163,35 +8553,38 @@
},
{
"id": "gpt-5-nano",
- "name": "GPT-5 Nano",
+ "name": "GPT-5 nano",
"provider": "openai",
- "family": "gpt5",
- "created_at": "2025-08-05 22:39:44 +0200",
- "context_window": 128000,
- "max_output_tokens": 400000,
+ "family": "gpt-5-nano",
+ "created_at": null,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.125
+ "input_per_million": 0.05,
+ "cached_input_per_million": 0.005,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.025,
+ "output_per_million": 0.2
}
}
},
@@ -8202,35 +8595,38 @@
},
{
"id": "gpt-5-nano-2025-08-07",
- "name": "GPT-5 Nano 20250807",
+ "name": "GPT-5 nano",
"provider": "openai",
- "family": "gpt5",
- "created_at": "2025-08-05 22:38:23 +0200",
- "context_window": 128000,
- "max_output_tokens": 400000,
+ "family": "gpt-5-nano",
+ "created_at": null,
+ "context_window": 400000,
+ "max_output_tokens": 128000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.125
+ "input_per_million": 0.05,
+ "cached_input_per_million": 0.005,
+ "output_per_million": 0.4
+ },
+ "batch": {
+ "input_per_million": 0.025,
+ "output_per_million": 0.2
}
}
},
@@ -8241,31 +8637,32 @@
},
{
"id": "gpt-audio",
- "name": "GPT-Audio",
+ "name": "gpt-audio",
"provider": "openai",
- "family": "other",
- "created_at": "2025-08-28 02:00:49 +0200",
- "context_window": 4096,
+ "family": "gpt-audio",
+ "created_at": null,
+ "context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "audio",
+ "text"
],
"output": [
- "text",
- "audio"
+ "audio",
+ "embeddings",
+ "text"
]
},
"capabilities": [
- "streaming"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -8276,31 +8673,32 @@
},
{
"id": "gpt-audio-2025-08-28",
- "name": "GPT-Audio 20250828",
+ "name": "gpt-audio",
"provider": "openai",
- "family": "other",
- "created_at": "2025-08-27 02:55:46 +0200",
- "context_window": 4096,
+ "family": "gpt-audio",
+ "created_at": null,
+ "context_window": 128000,
"max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "audio",
+ "text"
],
"output": [
- "text",
- "audio"
+ "audio",
+ "embeddings",
+ "text"
]
},
"capabilities": [
- "streaming"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
@@ -8311,30 +8709,30 @@
},
{
"id": "gpt-image-1",
- "name": "GPT-Image 1",
+ "name": "GPT Image 1",
"provider": "openai",
- "family": "other",
- "created_at": "2025-04-24 19:50:30 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "gpt-image-1",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
- "text",
+ "embeddings",
"image"
]
},
- "capabilities": [
- "streaming"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 5.0,
+ "cached_input_per_million": 1.25,
+ "output_per_million": 40.0
}
}
},
@@ -8343,31 +8741,88 @@
"owned_by": "system"
}
},
+ {
+ "id": "gpt-oss-120b",
+ "name": "gpt-oss-120b",
+ "provider": "openai",
+ "family": "gpt-oss-120b",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": 131072,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "text"
+ ]
+ },
+ "capabilities": [
+ "batch",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {},
+ "metadata": {}
+ },
+ {
+ "id": "gpt-oss-20b",
+ "name": "gpt-oss-20b",
+ "provider": "openai",
+ "family": "gpt-oss-20b",
+ "created_at": null,
+ "context_window": 131072,
+ "max_output_tokens": 131072,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "text"
+ ]
+ },
+ "capabilities": [
+ "batch",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {},
+ "metadata": {}
+ },
{
"id": "gpt-realtime",
- "name": "GPT-Realtime",
+ "name": "gpt-realtime",
"provider": "openai",
- "family": "other",
- "created_at": "2025-08-27 07:15:01 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "gpt-realtime",
+ "created_at": null,
+ "context_window": 32000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
+ "image",
"text"
],
"output": [
+ "audio",
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 4.0,
+ "cached_input_per_million": 0.4,
+ "output_per_million": 16.0
}
}
},
@@ -8378,29 +8833,34 @@
},
{
"id": "gpt-realtime-2025-08-28",
- "name": "GPT-Realtime 20250828",
+ "name": "gpt-realtime",
"provider": "openai",
- "family": "other",
- "created_at": "2025-08-27 07:16:13 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "gpt-realtime",
+ "created_at": null,
+ "context_window": 32000,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "audio",
+ "image",
"text"
],
"output": [
+ "audio",
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming"
+ "function_calling"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 4.0,
+ "cached_input_per_million": 0.4,
+ "output_per_million": 16.0
}
}
},
@@ -8411,37 +8871,41 @@
},
{
"id": "o1",
- "name": "O1",
+ "name": "o1",
"provider": "openai",
"family": "o1",
- "created_at": "2024-12-16 20:03:36 +0100",
+ "created_at": null,
"context_window": 200000,
"max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 15.0,
+ "cached_input_per_million": 7.5,
"output_per_million": 60.0
- }
- }
- },
+ },
+ "batch": {
+ "input_per_million": 7.5,
+ "output_per_million": 30.0
+ }
+ }
+ },
"metadata": {
"object": "model",
"owned_by": "system"
@@ -8449,34 +8913,38 @@
},
{
"id": "o1-2024-12-17",
- "name": "O1-20241217",
+ "name": "o1",
"provider": "openai",
"family": "o1",
- "created_at": "2024-12-16 06:29:36 +0100",
+ "created_at": null,
"context_window": 200000,
"max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 15.0,
+ "cached_input_per_million": 7.5,
"output_per_million": 60.0
+ },
+ "batch": {
+ "input_per_million": 7.5,
+ "output_per_million": 30.0
}
}
},
@@ -8487,10 +8955,10 @@
},
{
"id": "o1-mini",
- "name": "O1-Mini",
+ "name": "o1-mini",
"provider": "openai",
- "family": "o1_mini",
- "created_at": "2024-09-06 20:56:48 +0200",
+ "family": "o1-mini",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 65536,
"knowledge_cutoff": null,
@@ -8499,17 +8967,16 @@
"text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming",
- "reasoning"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 1.1,
+ "cached_input_per_million": 0.55,
"output_per_million": 4.4
}
}
@@ -8521,10 +8988,10 @@
},
{
"id": "o1-mini-2024-09-12",
- "name": "O1-Mini 20240912",
+ "name": "o1-mini",
"provider": "openai",
- "family": "o1_mini",
- "created_at": "2024-09-06 20:56:19 +0200",
+ "family": "o1-mini",
+ "created_at": null,
"context_window": 128000,
"max_output_tokens": 65536,
"knowledge_cutoff": null,
@@ -8533,17 +9000,16 @@
"text"
],
"output": [
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming",
- "reasoning"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 1.1,
+ "cached_input_per_million": 0.55,
"output_per_million": 4.4
}
}
@@ -8553,36 +9019,105 @@
"owned_by": "system"
}
},
+ {
+ "id": "o1-preview",
+ "name": "o1 Preview",
+ "provider": "openai",
+ "family": "o1-preview",
+ "created_at": null,
+ "context_window": 128000,
+ "max_output_tokens": 32768,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "text"
+ ]
+ },
+ "capabilities": [
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 15.0,
+ "cached_input_per_million": 7.5,
+ "output_per_million": 60.0
+ }
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "id": "o1-preview-2024-09-12",
+ "name": "o1 Preview",
+ "provider": "openai",
+ "family": "o1-preview",
+ "created_at": null,
+ "context_window": 128000,
+ "max_output_tokens": 32768,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "text"
+ ]
+ },
+ "capabilities": [
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 15.0,
+ "cached_input_per_million": 7.5,
+ "output_per_million": 60.0
+ }
+ }
+ },
+ "metadata": {}
+ },
{
"id": "o1-pro",
- "name": "O1-Pro",
+ "name": "o1-pro",
"provider": "openai",
- "family": "o1_pro",
- "created_at": "2025-03-17 23:49:51 +0100",
+ "family": "o1-pro",
+ "created_at": null,
"context_window": 200000,
"max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 150.0,
"output_per_million": 600.0
+ },
+ "batch": {
+ "input_per_million": 75.0,
+ "output_per_million": 300.0
}
}
},
@@ -8593,34 +9128,37 @@
},
{
"id": "o1-pro-2025-03-19",
- "name": "O1-Pro 20250319",
+ "name": "o1-pro",
"provider": "openai",
- "family": "o1_pro",
- "created_at": "2025-03-17 23:45:04 +0100",
+ "family": "o1-pro",
+ "created_at": null,
"context_window": 200000,
"max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 150.0,
"output_per_million": 600.0
+ },
+ "batch": {
+ "input_per_million": 75.0,
+ "output_per_million": 300.0
}
}
},
@@ -8631,30 +9169,38 @@
},
{
"id": "o3",
- "name": "O3",
+ "name": "o3",
"provider": "openai",
- "family": "other",
- "created_at": "2025-04-09 21:01:48 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o3",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
+ },
+ "batch": {
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
@@ -8665,30 +9211,38 @@
},
{
"id": "o3-2025-04-16",
- "name": "O3-20250416",
+ "name": "o3",
"provider": "openai",
- "family": "other",
- "created_at": "2025-04-08 19:28:21 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o3",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
+ },
+ "batch": {
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
@@ -8699,30 +9253,36 @@
},
{
"id": "o3-deep-research",
- "name": "O3-Deep Research",
+ "name": "o3-deep-research",
"provider": "openai",
- "family": "other",
- "created_at": "2025-06-13 20:42:01 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o3-deep-research",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 10.0,
+ "cached_input_per_million": 2.5,
+ "output_per_million": 40.0
+ },
+ "batch": {
+ "input_per_million": 5.0,
+ "output_per_million": 20.0
}
}
},
@@ -8733,30 +9293,36 @@
},
{
"id": "o3-deep-research-2025-06-26",
- "name": "O3-Deep Research 20250626",
+ "name": "o3-deep-research",
"provider": "openai",
- "family": "other",
- "created_at": "2025-06-25 17:26:59 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o3-deep-research",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 10.0,
+ "cached_input_per_million": 2.5,
+ "output_per_million": 40.0
+ },
+ "batch": {
+ "input_per_million": 5.0,
+ "output_per_million": 20.0
}
}
},
@@ -8767,10 +9333,10 @@
},
{
"id": "o3-mini",
- "name": "O3-Mini",
+ "name": "o3-mini",
"provider": "openai",
- "family": "o3_mini",
- "created_at": "2025-01-17 21:39:43 +0100",
+ "family": "o3-mini",
+ "created_at": null,
"context_window": 200000,
"max_output_tokens": 100000,
"knowledge_cutoff": null,
@@ -8779,20 +9345,25 @@
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 1.1,
+ "cached_input_per_million": 0.55,
"output_per_million": 4.4
+ },
+ "batch": {
+ "input_per_million": 0.55,
+ "output_per_million": 2.2
}
}
},
@@ -8803,10 +9374,10 @@
},
{
"id": "o3-mini-2025-01-31",
- "name": "O3-Mini 20250131",
+ "name": "o3-mini",
"provider": "openai",
- "family": "o3_mini",
- "created_at": "2025-01-27 21:36:40 +0100",
+ "family": "o3-mini",
+ "created_at": null,
"context_window": 200000,
"max_output_tokens": 100000,
"knowledge_cutoff": null,
@@ -8815,20 +9386,25 @@
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
+ "batch",
"function_calling",
- "structured_output",
- "reasoning"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 1.1,
+ "cached_input_per_million": 0.55,
"output_per_million": 4.4
+ },
+ "batch": {
+ "input_per_million": 0.55,
+ "output_per_million": 2.2
}
}
},
@@ -8839,30 +9415,37 @@
},
{
"id": "o3-pro",
- "name": "O3-Pro",
+ "name": "o3-pro",
"provider": "openai",
- "family": "other",
- "created_at": "2025-05-29 01:35:49 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o3-pro",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 20.0,
+ "output_per_million": 80.0
+ },
+ "batch": {
+ "input_per_million": 10.0,
+ "output_per_million": 40.0
}
}
},
@@ -8873,30 +9456,37 @@
},
{
"id": "o3-pro-2025-06-10",
- "name": "O3-Pro 20250610",
+ "name": "o3-pro",
"provider": "openai",
- "family": "other",
- "created_at": "2025-06-06 01:39:21 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o3-pro",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 20.0,
+ "output_per_million": 80.0
+ },
+ "batch": {
+ "input_per_million": 10.0,
+ "output_per_million": 40.0
}
}
},
@@ -8907,30 +9497,38 @@
},
{
"id": "o4-mini",
- "name": "O4 Mini",
+ "name": "o4-mini",
"provider": "openai",
- "family": "other",
- "created_at": "2025-04-09 21:02:31 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o4-mini",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 1.1,
+ "cached_input_per_million": 0.275,
+ "output_per_million": 4.4
+ },
+ "batch": {
+ "input_per_million": 0.55,
+ "output_per_million": 2.2
}
}
},
@@ -8941,30 +9539,38 @@
},
{
"id": "o4-mini-2025-04-16",
- "name": "O4 Mini 20250416",
+ "name": "o4-mini",
"provider": "openai",
- "family": "other",
- "created_at": "2025-04-08 19:31:46 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o4-mini",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 1.1,
+ "cached_input_per_million": 0.275,
+ "output_per_million": 4.4
+ },
+ "batch": {
+ "input_per_million": 0.55,
+ "output_per_million": 2.2
}
}
},
@@ -8975,30 +9581,36 @@
},
{
"id": "o4-mini-deep-research",
- "name": "O4 Mini Deep Research",
+ "name": "o4-mini-deep-research",
"provider": "openai",
- "family": "other",
- "created_at": "2025-06-12 01:44:45 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o4-mini-deep-research",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
+ },
+ "batch": {
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
@@ -9009,30 +9621,36 @@
},
{
"id": "o4-mini-deep-research-2025-06-26",
- "name": "O4 Mini Deep Research 20250626",
+ "name": "o4-mini-deep-research",
"provider": "openai",
- "family": "other",
- "created_at": "2025-06-25 17:42:01 +0200",
- "context_window": 4096,
- "max_output_tokens": 16384,
+ "family": "o4-mini-deep-research",
+ "created_at": null,
+ "context_window": 200000,
+ "max_output_tokens": 100000,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
+ "embeddings",
"text"
]
},
"capabilities": [
- "streaming",
- "reasoning"
+ "batch"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.0,
+ "cached_input_per_million": 0.5,
+ "output_per_million": 8.0
+ },
+ "batch": {
+ "input_per_million": 1.0,
+ "output_per_million": 4.0
}
}
},
@@ -9070,22 +9688,21 @@
},
{
"id": "omni-moderation-latest",
- "name": "Omni Moderation Latest",
+ "name": "omni-moderation",
"provider": "openai",
- "family": "moderation",
- "created_at": "2024-11-15 17:47:45 +0100",
+ "family": "omni-moderation-latest",
+ "created_at": null,
"context_window": null,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "pdf"
+ "text"
],
"output": [
- "text",
- "moderation"
+ "embeddings",
+ "text"
]
},
"capabilities": [],
@@ -9097,10 +9714,10 @@
},
{
"id": "text-embedding-3-large",
- "name": "text-embedding- 3 Large",
+ "name": "text-embedding-3-large",
"provider": "openai",
- "family": "embedding3_large",
- "created_at": "2024-01-22 20:53:00 +0100",
+ "family": "text-embedding-3-large",
+ "created_at": null,
"context_window": null,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -9109,8 +9726,8 @@
"text"
],
"output": [
- "text",
- "embeddings"
+ "embeddings",
+ "text"
]
},
"capabilities": [
@@ -9119,12 +9736,18 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.13,
- "output_per_million": 0.13
+ "input_per_million": 0.13
+ },
+ "batch": {
+ "input_per_million": 0.065
+ }
+ },
+ "embeddings": {
+ "standard": {
+ "input_per_million": 0.13
},
"batch": {
- "input_per_million": 0.065,
- "output_per_million": 0.065
+ "input_per_million": 0.065
}
}
},
@@ -9135,10 +9758,10 @@
},
{
"id": "text-embedding-3-small",
- "name": "text-embedding- 3 Small",
+ "name": "text-embedding-3-small",
"provider": "openai",
- "family": "embedding3_small",
- "created_at": "2024-01-22 19:43:17 +0100",
+ "family": "text-embedding-3-small",
+ "created_at": null,
"context_window": null,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -9147,8 +9770,8 @@
"text"
],
"output": [
- "text",
- "embeddings"
+ "embeddings",
+ "text"
]
},
"capabilities": [
@@ -9157,12 +9780,18 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.02,
- "output_per_million": 0.02
+ "input_per_million": 0.02
},
"batch": {
- "input_per_million": 0.01,
- "output_per_million": 0.01
+ "input_per_million": 0.01
+ }
+ },
+ "embeddings": {
+ "standard": {
+ "input_per_million": 0.02
+ },
+ "batch": {
+ "input_per_million": 0.01
}
}
},
@@ -9173,10 +9802,10 @@
},
{
"id": "text-embedding-ada-002",
- "name": "text-embedding- Ada 002",
+ "name": "text-embedding-ada-002",
"provider": "openai",
- "family": "embedding_ada",
- "created_at": "2022-12-16 20:01:39 +0100",
+ "family": "text-embedding-ada-002",
+ "created_at": null,
"context_window": null,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -9185,8 +9814,8 @@
"text"
],
"output": [
- "text",
- "embeddings"
+ "embeddings",
+ "text"
]
},
"capabilities": [
@@ -9195,12 +9824,18 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.1,
- "output_per_million": 0.1
+ "input_per_million": 0.1
},
"batch": {
- "input_per_million": 0.05,
- "output_per_million": 0.05
+ "input_per_million": 0.05
+ }
+ },
+ "embeddings": {
+ "standard": {
+ "input_per_million": 0.1
+ },
+ "batch": {
+ "input_per_million": 0.05
}
}
},
@@ -9209,33 +9844,73 @@
"owned_by": "openai-internal"
}
},
+ {
+ "id": "text-moderation-latest",
+ "name": "text-moderation",
+ "provider": "openai",
+ "family": "text-moderation-latest",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": 32768,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "text"
+ ]
+ },
+ "capabilities": [],
+ "pricing": {},
+ "metadata": {}
+ },
+ {
+ "id": "text-moderation-stable",
+ "name": "text-moderation-stable",
+ "provider": "openai",
+ "family": "text-moderation-stable",
+ "created_at": null,
+ "context_window": null,
+ "max_output_tokens": 32768,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "embeddings",
+ "text"
+ ]
+ },
+ "capabilities": [],
+ "pricing": {},
+ "metadata": {}
+ },
{
"id": "tts-1",
"name": "TTS-1",
"provider": "openai",
- "family": "tts1",
- "created_at": "2023-04-19 23:49:11 +0200",
+ "family": "tts-1",
+ "created_at": null,
"context_window": null,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "text"
],
"output": [
- "text",
- "audio"
+ "audio",
+ "embeddings"
]
},
- "capabilities": [
- "streaming"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 15.0,
- "output_per_million": 15.0
+ "input_per_million": 15.0
}
}
},
@@ -9283,29 +9958,25 @@
"id": "tts-1-hd",
"name": "TTS-1 HD",
"provider": "openai",
- "family": "tts1_hd",
- "created_at": "2023-11-03 22:13:35 +0100",
+ "family": "tts-1-hd",
+ "created_at": null,
"context_window": null,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "audio"
+ "text"
],
"output": [
- "text",
- "audio"
+ "audio",
+ "embeddings"
]
},
- "capabilities": [
- "streaming"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 30.0,
- "output_per_million": 30.0
+ "input_per_million": 30.0
}
}
},
@@ -9351,30 +10022,28 @@
},
{
"id": "whisper-1",
- "name": "Whisper 1",
+ "name": "Whisper",
"provider": "openai",
- "family": "whisper",
- "created_at": "2023-02-27 22:13:04 +0100",
+ "family": "whisper-1",
+ "created_at": null,
"context_window": null,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"audio"
],
"output": [
+ "audio",
+ "embeddings",
"text"
]
},
- "capabilities": [
- "streaming"
- ],
+ "capabilities": [],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.006,
- "output_per_million": 0.006
+ "input_per_million": 0.006
}
}
},
@@ -10379,7 +11048,7 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.0,
+ "input_per_million": 2.5,
"output_per_million": 5.0
}
}
@@ -12270,13 +12939,13 @@
}
},
{
- "id": "cohere/command",
- "name": "Cohere: Command",
+ "id": "cohere/command-a",
+ "name": "Cohere: Command A",
"provider": "openrouter",
"family": "cohere",
- "created_at": "2024-03-14 01:00:00 +0100",
- "context_window": 4096,
- "max_output_tokens": 4000,
+ "created_at": "2025-03-13 20:32:22 +0100",
+ "context_window": 256000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12293,13 +12962,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.0,
- "output_per_million": 2.0
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.\nCompared to other leading proprietary and open-weights models Command A delivers maximum performance with minimum hardware costs, excelling on business-critical agentic and multilingual tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12308,12 +12977,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
+ "tokenizer": "Other",
"instruct_type": null
},
"top_provider": {
- "context_length": 4096,
- "max_completion_tokens": 4000,
+ "context_length": 256000,
+ "max_completion_tokens": 8192,
"is_moderated": true
},
"per_request_limits": null,
@@ -12332,13 +13001,13 @@
}
},
{
- "id": "cohere/command-a",
- "name": "Cohere: Command A",
+ "id": "cohere/command-r-08-2024",
+ "name": "Cohere: Command R (08-2024)",
"provider": "openrouter",
"family": "cohere",
- "created_at": "2025-03-13 20:32:22 +0100",
- "context_window": 256000,
- "max_output_tokens": 8192,
+ "created_at": "2024-08-30 02:00:00 +0200",
+ "context_window": 128000,
+ "max_output_tokens": 4000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12350,18 +13019,19 @@
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 0.15,
+ "output_per_million": 0.6
}
}
},
"metadata": {
- "description": "Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.\nCompared to other leading proprietary and open-weights models Command A delivers maximum performance with minimum hardware costs, excelling on business-critical agentic and multilingual tasks.",
+ "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12370,12 +13040,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Other",
+ "tokenizer": "Cohere",
"instruct_type": null
},
"top_provider": {
- "context_length": 256000,
- "max_completion_tokens": 8192,
+ "context_length": 128000,
+ "max_completion_tokens": 4000,
"is_moderated": true
},
"per_request_limits": null,
@@ -12388,17 +13058,18 @@
"stop",
"structured_outputs",
"temperature",
+ "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "cohere/command-r",
- "name": "Cohere: Command R",
+ "id": "cohere/command-r-plus-08-2024",
+ "name": "Cohere: Command R+ (08-2024)",
"provider": "openrouter",
"family": "cohere",
- "created_at": "2024-03-14 01:00:00 +0100",
+ "created_at": "2024-08-30 02:00:00 +0200",
"context_window": 128000,
"max_output_tokens": 4000,
"knowledge_cutoff": null,
@@ -12418,13 +13089,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 2.5,
+ "output_per_million": 10.0
}
}
},
"metadata": {
- "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12458,11 +13129,11 @@
}
},
{
- "id": "cohere/command-r-03-2024",
- "name": "Cohere: Command R (03-2024)",
+ "id": "cohere/command-r7b-12-2024",
+ "name": "Cohere: Command R7B (12-2024)",
"provider": "openrouter",
"family": "cohere",
- "created_at": "2024-03-02 02:00:00 +0100",
+ "created_at": "2024-12-14 07:35:52 +0100",
"context_window": 128000,
"max_output_tokens": 4000,
"knowledge_cutoff": null,
@@ -12476,19 +13147,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.5,
- "output_per_million": 1.5
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
"metadata": {
- "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12515,20 +13185,19 @@
"stop",
"structured_outputs",
"temperature",
- "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "cohere/command-r-08-2024",
- "name": "Cohere: Command R (08-2024)",
+ "id": "deepcogito/cogito-v2-preview-deepseek-671b",
+ "name": "Deep Cogito: Cogito V2 Preview Deepseek 671B",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-08-30 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepcogito",
+ "created_at": "2025-09-02 18:35:49 +0200",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12540,19 +13209,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.6
+ "input_per_million": 1.25,
+ "output_per_million": 1.25
}
}
},
"metadata": {
- "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "Cogito v2 is a multilingual, instruction-tuned Mixture of Experts (MoE) large language model with 671 billion parameters. It supports both standard and reasoning-based generation modes. The model introduces hybrid reasoning via Iterated Distillation and Amplification (IDA)—an iterative self-improvement strategy designed to scale alignment with general intelligence. Cogito v2 has been optimized for STEM, programming, instruction following, and tool use. It supports 128k context length and offers strong performance in both multilingual and code-heavy environments. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12561,41 +13229,43 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
+ "tokenizer": "DeepSeek",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 163840,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
- "seed",
+ "reasoning",
+ "repetition_penalty",
"stop",
- "structured_outputs",
"temperature",
- "tools",
"top_k",
"top_p"
]
}
},
{
- "id": "cohere/command-r-plus",
- "name": "Cohere: Command R+",
+ "id": "deepcogito/cogito-v2-preview-llama-109b-moe",
+ "name": "Cogito V2 Preview Llama 109B",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-04-04 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepcogito",
+ "created_at": "2025-09-02 18:46:08 +0200",
+ "context_window": 32767,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "image",
"text"
],
"output": [
@@ -12605,44 +13275,48 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 0.18,
+ "output_per_million": 0.59
}
}
},
"metadata": {
- "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "An instruction-tuned, hybrid-reasoning Mixture-of-Experts model built on Llama-4-Scout-17B-16E. Cogito v2 can answer directly or engage an extended “thinking” phase, with alignment guided by Iterated Distillation & Amplification (IDA). It targets coding, STEM, instruction following, and general helpfulness, with stronger multilingual, tool-calling, and reasoning performance than size-equivalent baselines. The model supports long-context use (up to 10M tokens) and standard Transformers workflows. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
+ "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
+ "tokenizer": "Llama4",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 32767,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
- "seed",
+ "reasoning",
+ "repetition_penalty",
"stop",
- "structured_outputs",
"temperature",
+ "tool_choice",
"tools",
"top_k",
"top_p"
@@ -12650,13 +13324,13 @@
}
},
{
- "id": "cohere/command-r-plus-04-2024",
- "name": "Cohere: Command R+ (04-2024)",
+ "id": "deepseek/deepseek-chat",
+ "name": "DeepSeek: DeepSeek V3",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-04-02 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2024-12-26 20:28:40 +0100",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12669,18 +13343,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 3.0,
- "output_per_million": 15.0
+ "input_per_million": 0.24999987999999998,
+ "output_per_million": 0.999999888
}
}
},
"metadata": {
- "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12689,38 +13364,44 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
+ "tokenizer": "DeepSeek",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 163840,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
"tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "cohere/command-r-plus-08-2024",
- "name": "Cohere: Command R+ (08-2024)",
+ "id": "deepseek/deepseek-chat-v3-0324",
+ "name": "DeepSeek: DeepSeek V3 0324",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-08-30 02:00:00 +0200",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-03-24 14:59:15 +0100",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12733,18 +13414,19 @@
"capabilities": [
"streaming",
"function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 2.5,
- "output_per_million": 10.0
+ "input_per_million": 0.24999987999999998,
+ "output_per_million": 0.999999888
}
}
},
"metadata": {
- "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12753,38 +13435,44 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
+ "tokenizer": "DeepSeek",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 163840,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
"tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "cohere/command-r7b-12-2024",
- "name": "Cohere: Command R7B (12-2024)",
+ "id": "deepseek/deepseek-chat-v3-0324:free",
+ "name": "DeepSeek: DeepSeek V3 0324 (free)",
"provider": "openrouter",
- "family": "cohere",
- "created_at": "2024-12-14 07:35:52 +0100",
- "context_window": 128000,
- "max_output_tokens": 4000,
+ "family": "deepseek",
+ "created_at": "2025-03-24 14:59:15 +0100",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -12796,18 +13484,12 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).",
+ "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12816,35 +13498,40 @@
"output_modalities": [
"text"
],
- "tokenizer": "Cohere",
+ "tokenizer": "DeepSeek",
"instruct_type": null
},
"top_provider": {
- "context_length": 128000,
- "max_completion_tokens": 4000,
- "is_moderated": true
+ "context_length": 163840,
+ "max_completion_tokens": null,
+ "is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepcogito/cogito-v2-preview-deepseek-671b",
- "name": "Deep Cogito: Cogito V2 Preview Deepseek 671B",
+ "id": "deepseek/deepseek-chat-v3.1",
+ "name": "DeepSeek: DeepSeek V3.1",
"provider": "openrouter",
- "family": "deepcogito",
- "created_at": "2025-09-02 18:35:49 +0200",
+ "family": "deepseek",
+ "created_at": "2025-08-21 14:33:48 +0200",
"context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -12858,18 +13545,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 1.25
+ "input_per_million": 0.24999987999999998,
+ "output_per_million": 0.999999888
}
}
},
"metadata": {
- "description": "Cogito v2 is a multilingual, instruction-tuned Mixture of Experts (MoE) large language model with 671 billion parameters. It supports both standard and reasoning-based generation modes. The model introduces hybrid reasoning via Iterated Distillation and Amplification (IDA)—an iterative self-improvement strategy designed to scale alignment with general intelligence. Cogito v2 has been optimized for STEM, programming, instruction following, and tool use. It supports 128k context length and offers strong performance in both multilingual and code-heavy environments. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
+ "description": "DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)\n\nThe model improves tool use, code generation, and reasoning efficiency, achieving performance comparable to DeepSeek-R1 on difficult benchmarks while responding more quickly. It supports structured tool calling, code agents, and search agents, making it suitable for research, coding, and agentic workflows. \n\nIt succeeds the [DeepSeek V3-0324](/deepseek/deepseek-chat-v3-0324) model and performs well on a variety of tasks.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -12879,7 +13568,7 @@
"text"
],
"tokenizer": "DeepSeek",
- "instruct_type": null
+ "instruct_type": "deepseek-v3.1"
},
"top_provider": {
"context_length": 163840,
@@ -12891,30 +13580,36 @@
"frequency_penalty",
"include_reasoning",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"reasoning",
"repetition_penalty",
+ "response_format",
+ "seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepcogito/cogito-v2-preview-llama-109b-moe",
- "name": "Cogito V2 Preview Llama 109B",
+ "id": "deepseek/deepseek-chat-v3.1:free",
+ "name": "DeepSeek: DeepSeek V3.1 (free)",
"provider": "openrouter",
- "family": "deepcogito",
- "created_at": "2025-09-02 18:46:08 +0200",
- "context_window": 32767,
+ "family": "deepseek",
+ "created_at": "2025-08-21 14:33:48 +0200",
+ "context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "image",
"text"
],
"output": [
@@ -12924,32 +13619,24 @@
"capabilities": [
"streaming",
"function_calling",
- "predicted_outputs"
+ "structured_output"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.18,
- "output_per_million": 0.59
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "An instruction-tuned, hybrid-reasoning Mixture-of-Experts model built on Llama-4-Scout-17B-16E. Cogito v2 can answer directly or engage an extended “thinking” phase, with alignment guided by Iterated Distillation & Amplification (IDA). It targets coding, STEM, instruction following, and general helpfulness, with stronger multilingual, tool-calling, and reasoning performance than size-equivalent baselines. The model supports long-context use (up to 10M tokens) and standard Transformers workflows. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)",
+ "description": "DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)\n\nThe model improves tool use, code generation, and reasoning efficiency, achieving performance comparable to DeepSeek-R1 on difficult benchmarks while responding more quickly. It supports structured tool calling, code agents, and search agents, making it suitable for research, coding, and agentic workflows. \n\nIt succeeds the [DeepSeek V3-0324](/deepseek/deepseek-chat-v3-0324) model and performs well on a variety of tasks.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "image",
"text"
],
"output_modalities": [
"text"
],
- "tokenizer": "Llama4",
- "instruct_type": null
+ "tokenizer": "DeepSeek",
+ "instruct_type": "deepseek-v3.1"
},
"top_provider": {
- "context_length": 32767,
+ "context_length": 163840,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -12957,12 +13644,13 @@
"supported_parameters": [
"frequency_penalty",
"include_reasoning",
- "logit_bias",
"max_tokens",
"min_p",
"presence_penalty",
"reasoning",
"repetition_penalty",
+ "response_format",
+ "seed",
"stop",
"temperature",
"tool_choice",
@@ -12973,11 +13661,11 @@
}
},
{
- "id": "deepseek/deepseek-chat",
- "name": "DeepSeek: DeepSeek V3",
+ "id": "deepseek/deepseek-prover-v2",
+ "name": "DeepSeek: DeepSeek Prover V2",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2024-12-26 20:28:40 +0100",
+ "created_at": "2025-04-30 13:38:14 +0200",
"context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -12991,20 +13679,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
- "predicted_outputs"
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.24999987999999998,
- "output_per_million": 0.999999888
+ "input_per_million": 0.5,
+ "output_per_million": 2.1799999999999997
}
}
},
"metadata": {
- "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).",
+ "description": "DeepSeek Prover V2 is a 671B parameter model, speculated to be geared towards logic and mathematics. Likely an upgrade from [DeepSeek-Prover-V1.5](https://huggingface.co/deepseek-ai/DeepSeek-Prover-V1.5-RL) Not much is known about the model yet, as DeepSeek released it on Hugging Face without an announcement or description.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13024,8 +13710,6 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -13033,24 +13717,20 @@
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-chat-v3-0324",
- "name": "DeepSeek: DeepSeek V3 0324",
+ "id": "deepseek/deepseek-r1",
+ "name": "DeepSeek: R1",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-03-24 14:59:15 +0100",
+ "created_at": "2025-01-20 14:51:35 +0100",
"context_window": 163840,
- "max_output_tokens": null,
+ "max_output_tokens": 163840,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13069,13 +13749,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.24999987999999998,
- "output_per_million": 0.999999888
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 2.0
}
}
},
"metadata": {
- "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.",
+ "description": "DeepSeek R1 is here: Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model & [technical report](https://api-docs.deepseek.com/news/news250120).\n\nMIT licensed: Distill & commercialize freely!",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13085,21 +13765,23 @@
"text"
],
"tokenizer": "DeepSeek",
- "instruct_type": null
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
"context_length": 163840,
- "max_completion_tokens": null,
+ "max_completion_tokens": 163840,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
"response_format",
"seed",
@@ -13115,11 +13797,11 @@
}
},
{
- "id": "deepseek/deepseek-chat-v3-0324:free",
- "name": "DeepSeek: DeepSeek V3 0324 (free)",
+ "id": "deepseek/deepseek-r1-0528",
+ "name": "DeepSeek: R1 0528",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-03-24 14:59:15 +0100",
+ "created_at": "2025-05-28 19:59:30 +0200",
"context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -13134,11 +13816,19 @@
"capabilities": [
"streaming",
"function_calling",
+ "structured_output",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 1.75
+ }
+ }
+ },
"metadata": {
- "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.",
+ "description": "May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13148,7 +13838,7 @@
"text"
],
"tokenizer": "DeepSeek",
- "instruct_type": null
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
"context_length": 163840,
@@ -13158,14 +13848,18 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -13176,12 +13870,12 @@
}
},
{
- "id": "deepseek/deepseek-chat-v3.1",
- "name": "DeepSeek: DeepSeek V3.1",
+ "id": "deepseek/deepseek-r1-0528-qwen3-8b",
+ "name": "DeepSeek: Deepseek R1 0528 Qwen3 8B",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-08-21 14:33:48 +0200",
- "context_window": 163840,
+ "created_at": "2025-05-29 19:09:03 +0200",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -13194,20 +13888,18 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.24999987999999998,
- "output_per_million": 0.999999888
+ "input_per_million": 0.01,
+ "output_per_million": 0.049999999999999996
}
}
},
"metadata": {
- "description": "DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)\n\nThe model improves tool use, code generation, and reasoning efficiency, achieving performance comparable to DeepSeek-R1 on difficult benchmarks while responding more quickly. It supports structured tool calling, code agents, and search agents, making it suitable for research, coding, and agentic workflows. \n\nIt succeeds the [DeepSeek V3-0324](/deepseek/deepseek-chat-v3-0324) model and performs well on a variety of tasks.",
+ "description": "DeepSeek-R1-0528 is a lightly upgraded release of DeepSeek R1 that taps more compute and smarter post-training tricks, pushing its reasoning and inference to the brink of flagship models like O3 and Gemini 2.5 Pro.\nIt now tops math, programming, and logic leaderboards, showcasing a step-change in depth-of-thought.\nThe distilled variant, DeepSeek-R1-0528-Qwen3-8B, transfers this chain-of-thought into an 8 B-parameter form, beating standard Qwen3 8B by +10 pp and tying the 235 B “thinking” giant on AIME 2024.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13216,11 +13908,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "deepseek-v3.1"
+ "tokenizer": "Qwen",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -13235,13 +13927,9 @@
"presence_penalty",
"reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -13249,12 +13937,12 @@
}
},
{
- "id": "deepseek/deepseek-chat-v3.1:free",
- "name": "DeepSeek: DeepSeek V3.1 (free)",
+ "id": "deepseek/deepseek-r1-0528-qwen3-8b:free",
+ "name": "DeepSeek: Deepseek R1 0528 Qwen3 8B (free)",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-08-21 14:33:48 +0200",
- "context_window": 163840,
+ "created_at": "2025-05-29 19:09:03 +0200",
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -13267,12 +13955,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
"pricing": {},
"metadata": {
- "description": "DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)\n\nThe model improves tool use, code generation, and reasoning efficiency, achieving performance comparable to DeepSeek-R1 on difficult benchmarks while responding more quickly. It supports structured tool calling, code agents, and search agents, making it suitable for research, coding, and agentic workflows. \n\nIt succeeds the [DeepSeek V3-0324](/deepseek/deepseek-chat-v3-0324) model and performs well on a variety of tasks.",
+ "description": "DeepSeek-R1-0528 is a lightly upgraded release of DeepSeek R1 that taps more compute and smarter post-training tricks, pushing its reasoning and inference to the brink of flagship models like O3 and Gemini 2.5 Pro.\nIt now tops math, programming, and logic leaderboards, showcasing a step-change in depth-of-thought.\nThe distilled variant, DeepSeek-R1-0528-Qwen3-8B, transfers this chain-of-thought into an 8 B-parameter form, beating standard Qwen3 8B by +10 pp and tying the 235 B “thinking” giant on AIME 2024.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13281,11 +13968,11 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "deepseek-v3.1"
+ "tokenizer": "Qwen",
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
+ "context_length": 131072,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -13293,28 +13980,28 @@
"supported_parameters": [
"frequency_penalty",
"include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
"temperature",
- "tool_choice",
- "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-prover-v2",
- "name": "DeepSeek: DeepSeek Prover V2",
+ "id": "deepseek/deepseek-r1-0528:free",
+ "name": "DeepSeek: R1 0528 (free)",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-04-30 13:38:14 +0200",
+ "created_at": "2025-05-28 19:59:30 +0200",
"context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
@@ -13328,18 +14015,11 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.5,
- "output_per_million": 2.1799999999999997
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "DeepSeek Prover V2 is a 671B parameter model, speculated to be geared towards logic and mathematics. Likely an upgrade from [DeepSeek-Prover-V1.5](https://huggingface.co/deepseek-ai/DeepSeek-Prover-V1.5-RL) Not much is known about the model yet, as DeepSeek released it on Hugging Face without an announcement or description.",
+ "description": "May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13349,7 +14029,7 @@
"text"
],
"tokenizer": "DeepSeek",
- "instruct_type": null
+ "instruct_type": "deepseek-r1"
},
"top_provider": {
"context_length": 163840,
@@ -13359,27 +14039,31 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
"temperature",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1",
- "name": "DeepSeek: R1",
+ "id": "deepseek/deepseek-r1-distill-llama-70b",
+ "name": "DeepSeek: R1 Distill Llama 70B",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-01-20 14:51:35 +0100",
- "context_window": 163840,
- "max_output_tokens": 163840,
+ "created_at": "2025-01-23 21:12:49 +0100",
+ "context_window": 131072,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13398,13 +14082,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 2.0
+ "input_per_million": 0.03,
+ "output_per_million": 0.13
}
}
},
"metadata": {
- "description": "DeepSeek R1 is here: Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model & [technical report](https://api-docs.deepseek.com/news/news250120).\n\nMIT licensed: Distill & commercialize freely!",
+ "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13413,12 +14097,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
+ "tokenizer": "Llama3",
"instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": 163840,
+ "context_length": 131072,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -13435,7 +14119,6 @@
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -13446,13 +14129,13 @@
}
},
{
- "id": "deepseek/deepseek-r1-0528",
- "name": "DeepSeek: R1 0528",
+ "id": "deepseek/deepseek-r1-distill-llama-70b:free",
+ "name": "DeepSeek: R1 Distill Llama 70B (free)",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-05-28 19:59:30 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "created_at": "2025-01-23 21:12:49 +0100",
+ "context_window": 8192,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13464,20 +14147,11 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output",
"predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.39999999999999997,
- "output_per_million": 1.75
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model.",
+ "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13486,12 +14160,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
+ "tokenizer": "Llama3",
"instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 8192,
+ "max_completion_tokens": 4096,
"is_moderated": false
},
"per_request_limits": null,
@@ -13505,13 +14179,9 @@
"presence_penalty",
"reasoning",
"repetition_penalty",
- "response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_k",
"top_logprobs",
"top_p"
@@ -13519,13 +14189,13 @@
}
},
{
- "id": "deepseek/deepseek-r1-0528-qwen3-8b",
- "name": "DeepSeek: Deepseek R1 0528 Qwen3 8B",
+ "id": "deepseek/deepseek-r1-distill-llama-8b",
+ "name": "DeepSeek: R1 Distill Llama 8B",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-05-29 19:09:03 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "created_at": "2025-02-07 15:15:18 +0100",
+ "context_window": 32000,
+ "max_output_tokens": 32000,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13542,13 +14212,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01,
- "output_per_million": 0.049999999999999996
+ "input_per_million": 0.04,
+ "output_per_million": 0.04
}
}
},
"metadata": {
- "description": "DeepSeek-R1-0528 is a lightly upgraded release of DeepSeek R1 that taps more compute and smarter post-training tricks, pushing its reasoning and inference to the brink of flagship models like O3 and Gemini 2.5 Pro.\nIt now tops math, programming, and logic leaderboards, showcasing a step-change in depth-of-thought.\nThe distilled variant, DeepSeek-R1-0528-Qwen3-8B, transfers this chain-of-thought into an 8 B-parameter form, beating standard Qwen3 8B by +10 pp and tying the 235 B “thinking” giant on AIME 2024.",
+ "description": "DeepSeek R1 Distill Llama 8B is a distilled large language model based on [Llama-3.1-8B-Instruct](/meta-llama/llama-3.1-8b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 50.4\n- MATH-500 pass@1: 89.1\n- CodeForces Rating: 1205\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.\n\nHugging Face: \n- [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) \n- [DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13557,12 +14227,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
+ "tokenizer": "Llama3",
"instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
+ "context_length": 32000,
+ "max_completion_tokens": 32000,
"is_moderated": false
},
"per_request_limits": null,
@@ -13570,7 +14240,6 @@
"frequency_penalty",
"include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -13580,19 +14249,18 @@
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-0528-qwen3-8b:free",
- "name": "DeepSeek: Deepseek R1 0528 Qwen3 8B (free)",
+ "id": "deepseek/deepseek-r1-distill-qwen-14b",
+ "name": "DeepSeek: R1 Distill Qwen 14B",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-05-29 19:09:03 +0200",
- "context_window": 131072,
- "max_output_tokens": null,
+ "created_at": "2025-01-30 00:39:00 +0100",
+ "context_window": 32768,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13606,9 +14274,16 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.15,
+ "output_per_million": 0.15
+ }
+ }
+ },
"metadata": {
- "description": "DeepSeek-R1-0528 is a lightly upgraded release of DeepSeek R1 that taps more compute and smarter post-training tricks, pushing its reasoning and inference to the brink of flagship models like O3 and Gemini 2.5 Pro.\nIt now tops math, programming, and logic leaderboards, showcasing a step-change in depth-of-thought.\nThe distilled variant, DeepSeek-R1-0528-Qwen3-8B, transfers this chain-of-thought into an 8 B-parameter form, beating standard Qwen3 8B by +10 pp and tying the 235 B “thinking” giant on AIME 2024.",
+ "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13621,8 +14296,8 @@
"instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": null,
+ "context_length": 32768,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
@@ -13630,7 +14305,6 @@
"frequency_penalty",
"include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -13640,19 +14314,18 @@
"stop",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-0528:free",
- "name": "DeepSeek: R1 0528 (free)",
+ "id": "deepseek/deepseek-r1-distill-qwen-32b",
+ "name": "DeepSeek: R1 Distill Qwen 32B",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-05-28 19:59:30 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "created_at": "2025-01-30 00:53:50 +0100",
+ "context_window": 131072,
+ "max_output_tokens": 16384,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13664,11 +14337,19 @@
},
"capabilities": [
"streaming",
+ "structured_output",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.27,
+ "output_per_million": 0.27
+ }
+ }
+ },
"metadata": {
- "description": "May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model.",
+ "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\\n\\nOther benchmark results include:\\n\\n- AIME 2024 pass@1: 72.6\\n- MATH-500 pass@1: 94.3\\n- CodeForces Rating: 1691\\n\\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13677,12 +14358,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
+ "tokenizer": "Qwen",
"instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 16384,
"is_moderated": false
},
"per_request_limits": null,
@@ -13690,28 +14371,28 @@
"frequency_penalty",
"include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"reasoning",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-llama-70b",
- "name": "DeepSeek: R1 Distill Llama 70B",
+ "id": "deepseek/deepseek-r1:free",
+ "name": "DeepSeek: R1 (free)",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-01-23 21:12:49 +0100",
- "context_window": 131072,
+ "created_at": "2025-01-20 14:51:35 +0100",
+ "context_window": 163840,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -13723,21 +14404,11 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "structured_output",
- "predicted_outputs"
+ "streaming"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.03,
- "output_per_million": 0.13
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "DeepSeek R1 is here: Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model & [technical report](https://api-docs.deepseek.com/news/news250120).\n\nMIT licensed: Distill & commercialize freely!",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13746,45 +14417,31 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
+ "tokenizer": "DeepSeek",
"instruct_type": "deepseek-r1"
},
"top_provider": {
- "context_length": 131072,
+ "context_length": 163840,
"max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"include_reasoning",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
"reasoning",
- "repetition_penalty",
- "response_format",
- "seed",
- "stop",
- "temperature",
- "tool_choice",
- "tools",
- "top_k",
- "top_logprobs",
- "top_p"
+ "temperature"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-llama-70b:free",
- "name": "DeepSeek: R1 Distill Llama 70B (free)",
+ "id": "deepseek/deepseek-v3.1-base",
+ "name": "DeepSeek: DeepSeek V3.1 Base",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-01-23 21:12:49 +0100",
- "context_window": 8192,
- "max_output_tokens": 4096,
+ "created_at": "2025-08-20 23:56:57 +0200",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13798,9 +14455,16 @@
"streaming",
"predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.24999987999999998,
+ "output_per_million": 0.999999888
+ }
+ }
+ },
"metadata": {
- "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "This is a base model, trained only for raw next-token prediction. Unlike instruct/chat models, it has not been fine-tuned to follow user instructions. Prompts need to be written more like training text or examples rather than simple requests (e.g., “Translate the following sentence…” instead of just “Translate this”).\n\nDeepSeek-V3.1 Base is a 671B parameter open Mixture-of-Experts (MoE) language model with 37B active parameters per forward pass and a context length of 128K tokens. Trained on 14.8T tokens using FP8 mixed precision, it achieves high training efficiency and stability, with strong performance across language, reasoning, math, and coding tasks. \n",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13809,24 +14473,22 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "DeepSeek",
+ "instruct_type": "none"
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 4096,
+ "context_length": 163840,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
"logit_bias",
"logprobs",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
"seed",
"stop",
@@ -13838,13 +14500,13 @@
}
},
{
- "id": "deepseek/deepseek-r1-distill-llama-8b",
- "name": "DeepSeek: R1 Distill Llama 8B",
+ "id": "deepseek/deepseek-v3.1-terminus",
+ "name": "DeepSeek: DeepSeek V3.1 Terminus",
"provider": "openrouter",
"family": "deepseek",
- "created_at": "2025-02-07 15:15:18 +0100",
- "context_window": 32000,
- "max_output_tokens": 32000,
+ "created_at": "2025-09-22 15:37:55 +0200",
+ "context_window": 163840,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13856,18 +14518,20 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.04,
- "output_per_million": 0.04
+ "input_per_million": 0.27,
+ "output_per_million": 1.0
}
}
},
"metadata": {
- "description": "DeepSeek R1 Distill Llama 8B is a distilled large language model based on [Llama-3.1-8B-Instruct](/meta-llama/llama-3.1-8b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 50.4\n- MATH-500 pass@1: 89.1\n- CodeForces Rating: 1205\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.\n\nHugging Face: \n- [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) \n- [DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |",
+ "description": "DeepSeek-V3.1 Terminus is an update to [DeepSeek V3.1](/deepseek/deepseek-chat-v3.1) that maintains the model's original capabilities while addressing issues reported by users, including language consistency and agent capabilities, further optimizing the model's performance in coding and search agents. It is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes. It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)\n\nThe model improves tool use, code generation, and reasoning efficiency, achieving performance comparable to DeepSeek-R1 on difficult benchmarks while responding more quickly. It supports structured tool calling, code agents, and search agents, making it suitable for research, coding, and agentic workflows. ",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13876,12 +14540,12 @@
"output_modalities": [
"text"
],
- "tokenizer": "Llama3",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "DeepSeek",
+ "instruct_type": "deepseek-v3.1"
},
"top_provider": {
- "context_length": 32000,
- "max_completion_tokens": 32000,
+ "context_length": 163840,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -13889,27 +14553,32 @@
"frequency_penalty",
"include_reasoning",
"logit_bias",
+ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
"reasoning",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1-distill-qwen-14b",
- "name": "DeepSeek: R1 Distill Qwen 14B",
+ "id": "eleutherai/llemma_7b",
+ "name": "EleutherAI: Llemma 7b",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-30 00:39:00 +0100",
- "context_window": 32768,
- "max_output_tokens": 16384,
+ "family": "eleutherai",
+ "created_at": "2025-04-14 17:07:05 +0200",
+ "context_window": 4096,
+ "max_output_tokens": 4096,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -13920,19 +14589,18 @@
]
},
"capabilities": [
- "streaming",
- "predicted_outputs"
+ "streaming"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.15
+ "input_per_million": 0.7999999999999999,
+ "output_per_million": 1.2
}
}
},
"metadata": {
- "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "Llemma 7B is a language model for mathematics. It was initialized with Code Llama 7B weights, and trained on the Proof-Pile-2 for 200B tokens. Llemma models are particularly strong at chain-of-thought mathematical reasoning and using computational tools for mathematics, such as Python and formal theorem provers.",
"architecture": {
"modality": "text->text",
"input_modalities": [
@@ -13941,23 +14609,20 @@
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Other",
+ "instruct_type": "code-llama"
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 16384,
+ "context_length": 4096,
+ "max_completion_tokens": 4096,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
- "include_reasoning",
- "logit_bias",
"max_tokens",
"min_p",
"presence_penalty",
- "reasoning",
"repetition_penalty",
"seed",
"stop",
@@ -13968,17 +14633,20 @@
}
},
{
- "id": "deepseek/deepseek-r1-distill-qwen-32b",
- "name": "DeepSeek: R1 Distill Qwen 32B",
+ "id": "google/gemini-2.0-flash-001",
+ "name": "Google: Gemini 2.0 Flash",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-30 00:53:50 +0100",
- "context_window": 131072,
- "max_output_tokens": 16384,
+ "family": "google",
+ "created_at": "2025-02-05 16:30:13 +0100",
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output": [
"text"
@@ -13986,114 +14654,124 @@
},
"capabilities": [
"streaming",
- "structured_output",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.27,
- "output_per_million": 0.27
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.39999999999999997,
+ "cached_input_per_million": 0.024999999999999998
}
}
},
"metadata": {
- "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\\n\\nOther benchmark results include:\\n\\n- AIME 2024 pass@1: 72.6\\n- MATH-500 pass@1: 94.3\\n- CodeForces Rating: 1691\\n\\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.",
+ "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 131072,
- "max_completion_tokens": 16384,
+ "context_length": 1048576,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "include_reasoning",
- "logit_bias",
"max_tokens",
- "min_p",
- "presence_penalty",
- "reasoning",
- "repetition_penalty",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "deepseek/deepseek-r1:free",
- "name": "DeepSeek: R1 (free)",
+ "id": "google/gemini-2.0-flash-exp:free",
+ "name": "Google: Gemini 2.0 Flash Experimental (free)",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-01-20 14:51:35 +0100",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "google",
+ "created_at": "2024-12-11 18:18:43 +0100",
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {},
"metadata": {
- "description": "DeepSeek R1 is here: Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.\n\nFully open-source model & [technical report](https://api-docs.deepseek.com/news/news250120).\n\nMIT licensed: Distill & commercialize freely!",
+ "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
"max_tokens",
- "reasoning",
- "temperature"
+ "response_format",
+ "seed",
+ "stop",
+ "temperature",
+ "tool_choice",
+ "tools",
+ "top_p"
]
}
},
{
- "id": "deepseek/deepseek-v3.1-base",
- "name": "DeepSeek: DeepSeek V3.1 Base",
+ "id": "google/gemini-2.0-flash-lite-001",
+ "name": "Google: Gemini 2.0 Flash Lite",
"provider": "openrouter",
- "family": "deepseek",
- "created_at": "2025-08-20 23:56:57 +0200",
- "context_window": 163840,
- "max_output_tokens": null,
+ "family": "google",
+ "created_at": "2025-02-25 18:56:52 +0100",
+ "context_window": 1048576,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output": [
"text"
@@ -14101,165 +14779,171 @@
},
"capabilities": [
"streaming",
- "predicted_outputs"
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.24999987999999998,
- "output_per_million": 0.999999888
+ "input_per_million": 0.075,
+ "output_per_million": 0.3
}
}
},
"metadata": {
- "description": "This is a base model, trained only for raw next-token prediction. Unlike instruct/chat models, it has not been fine-tuned to follow user instructions. Prompts need to be written more like training text or examples rather than simple requests (e.g., “Translate the following sentence…” instead of just “Translate this”).\n\nDeepSeek-V3.1 Base is a 671B parameter open Mixture-of-Experts (MoE) language model with 37B active parameters per forward pass and a context length of 128K tokens. Trained on 14.8T tokens using FP8 mixed precision, it achieves high training efficiency and stability, with strong performance across language, reasoning, math, and coding tasks. \n",
+ "description": "Gemini 2.0 Flash Lite offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5), all at extremely economical token prices.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image",
+ "file",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "DeepSeek",
- "instruct_type": "none"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 163840,
- "max_completion_tokens": null,
+ "context_length": 1048576,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
- "top_logprobs",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "eleutherai/llemma_7b",
- "name": "EleutherAI: Llemma 7b",
+ "id": "google/gemini-2.5-flash",
+ "name": "Google: Gemini 2.5 Flash",
"provider": "openrouter",
- "family": "eleutherai",
- "created_at": "2025-04-14 17:07:05 +0200",
- "context_window": 4096,
- "max_output_tokens": 4096,
+ "family": "google",
+ "created_at": "2025-06-17 17:01:28 +0200",
+ "context_window": 1048576,
+ "max_output_tokens": 65535,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output": [
"text"
]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.7999999999999999,
- "output_per_million": 1.2
+ "input_per_million": 0.3,
+ "output_per_million": 2.5,
+ "cached_input_per_million": 0.075
}
}
},
"metadata": {
- "description": "Llemma 7B is a language model for mathematics. It was initialized with Code Llama 7B weights, and trained on the Proof-Pile-2 for 200B tokens. Llemma models are particularly strong at chain-of-thought mathematical reasoning and using computational tools for mathematics, such as Python and formal theorem provers.",
+ "description": "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning).",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "file",
+ "image",
+ "text",
+ "audio"
],
"output_modalities": [
"text"
],
- "tokenizer": "Other",
- "instruct_type": "code-llama"
+ "tokenizer": "Gemini",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 4096,
- "max_completion_tokens": 4096,
+ "context_length": 1048576,
+ "max_completion_tokens": 65535,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
+ "include_reasoning",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
+ "reasoning",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
- "top_k",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "google/gemini-2.0-flash-001",
- "name": "Google: Gemini 2.0 Flash",
+ "id": "google/gemini-2.5-flash-image-preview",
+ "name": "Gemini 2.5 Flash Image (Nano Banana)",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-02-05 16:30:13 +0100",
- "context_window": 1048576,
+ "created_at": "2025-08-26 16:36:17 +0200",
+ "context_window": 32768,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
"image",
- "file",
- "audio"
+ "text"
],
"output": [
+ "image",
"text"
]
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.39999999999999997,
- "cached_input_per_million": 0.024999999999999998
+ "input_per_million": 0.3,
+ "output_per_million": 2.5
}
}
},
"metadata": {
- "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.",
+ "description": "Gemini 2.5 Flash Image Preview, a.k.a. \"Nano Banana,\" is a state of the art image generation model with contextual understanding. It is capable of image generation, edits, and multi-turn conversations.",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text+image->text+image",
"input_modalities": [
- "text",
"image",
- "file",
- "audio"
+ "text"
],
"output_modalities": [
+ "image",
"text"
],
"tokenizer": "Gemini",
"instruct_type": null
},
"top_provider": {
- "context_length": 1048576,
+ "context_length": 32768,
"max_completion_tokens": 8192,
"is_moderated": false
},
@@ -14268,28 +14952,27 @@
"max_tokens",
"response_format",
"seed",
- "stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_p"
]
}
},
{
- "id": "google/gemini-2.0-flash-exp:free",
- "name": "Google: Gemini 2.0 Flash Experimental (free)",
+ "id": "google/gemini-2.5-flash-lite",
+ "name": "Google: Gemini 2.5 Flash Lite",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-12-11 18:18:43 +0100",
+ "created_at": "2025-07-22 18:04:36 +0200",
"context_window": 1048576,
- "max_output_tokens": 8192,
+ "max_output_tokens": 65535,
"knowledge_cutoff": null,
"modalities": {
"input": [
+ "file",
+ "image",
"text",
- "image"
+ "audio"
],
"output": [
"text"
@@ -14300,14 +14983,24 @@
"function_calling",
"structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.39999999999999997,
+ "cached_input_per_million": 0.024999999999999998
+ }
+ }
+ },
"metadata": {
- "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.",
+ "description": "Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, \"thinking\" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. ",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
+ "file",
+ "image",
"text",
- "image"
+ "audio"
],
"output_modalities": [
"text"
@@ -14317,15 +15010,18 @@
},
"top_provider": {
"context_length": 1048576,
- "max_completion_tokens": 8192,
+ "max_completion_tokens": 65535,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "include_reasoning",
"max_tokens",
+ "reasoning",
"response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
"tool_choice",
"tools",
@@ -14334,19 +15030,19 @@
}
},
{
- "id": "google/gemini-2.0-flash-lite-001",
- "name": "Google: Gemini 2.0 Flash Lite",
+ "id": "google/gemini-2.5-flash-lite-preview-06-17",
+ "name": "Google: Gemini 2.5 Flash Lite Preview 06-17",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-02-25 18:56:52 +0100",
+ "created_at": "2025-06-17 17:23:51 +0200",
"context_window": 1048576,
- "max_output_tokens": 8192,
+ "max_output_tokens": 65535,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
"file",
+ "image",
+ "text",
"audio"
],
"output": [
@@ -14361,19 +15057,20 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.39999999999999997,
+ "cached_input_per_million": 0.024999999999999998
}
}
},
"metadata": {
- "description": "Gemini 2.0 Flash Lite offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5), all at extremely economical token prices.",
+ "description": "Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, \"thinking\" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. ",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "text",
- "image",
"file",
+ "image",
+ "text",
"audio"
],
"output_modalities": [
@@ -14384,12 +15081,14 @@
},
"top_provider": {
"context_length": 1048576,
- "max_completion_tokens": 8192,
+ "max_completion_tokens": 65535,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "include_reasoning",
"max_tokens",
+ "reasoning",
"response_format",
"seed",
"stop",
@@ -14402,13 +15101,13 @@
}
},
{
- "id": "google/gemini-2.5-flash",
- "name": "Google: Gemini 2.5 Flash",
+ "id": "google/gemini-2.5-flash-lite-preview-09-2025",
+ "name": "Google: Gemini 2.5 Flash Lite Preview 09-2025",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-06-17 17:01:28 +0200",
+ "created_at": "2025-09-25 19:01:26 +0200",
"context_window": 1048576,
- "max_output_tokens": 65535,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -14429,14 +15128,13 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.3,
- "output_per_million": 2.5,
- "cached_input_per_million": 0.075
+ "input_per_million": 0.09999999999999999,
+ "output_per_million": 0.39999999999999997
}
}
},
"metadata": {
- "description": "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning).",
+ "description": "Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, \"thinking\" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. ",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -14453,7 +15151,7 @@
},
"top_provider": {
"context_length": 1048576,
- "max_completion_tokens": 65535,
+ "max_completion_tokens": 65536,
"is_moderated": false
},
"per_request_limits": null,
@@ -14473,75 +15171,82 @@
}
},
{
- "id": "google/gemini-2.5-flash-image-preview",
- "name": "Google: Gemini 2.5 Flash Image Preview",
+ "id": "google/gemini-2.5-flash-preview-09-2025",
+ "name": "Google: Gemini 2.5 Flash Preview 09-2025",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-08-26 16:36:17 +0200",
- "context_window": 32768,
- "max_output_tokens": 8192,
+ "created_at": "2025-09-25 19:09:38 +0200",
+ "context_window": 1048576,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
"image",
+ "file",
"text"
],
"output": [
- "image",
"text"
]
},
"capabilities": [
"streaming",
+ "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
"input_per_million": 0.3,
- "output_per_million": 2.5
+ "output_per_million": 2.5,
+ "cached_input_per_million": 0.075
}
}
},
"metadata": {
- "description": "Gemini 2.5 Flash Image Preview, AKA Nano Banana is a state of the art image generation model with contextual understanding. It is capable of image generation, edits, and multi-turn conversations.",
+ "description": "Gemini 2.5 Flash Preview September 2025 Checkpoint is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning).",
"architecture": {
- "modality": "text+image->text+image",
+ "modality": "text+image->text",
"input_modalities": [
"image",
+ "file",
"text"
],
"output_modalities": [
- "image",
"text"
],
"tokenizer": "Gemini",
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 8192,
+ "context_length": 1048576,
+ "max_completion_tokens": 65536,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
+ "include_reasoning",
"max_tokens",
+ "reasoning",
"response_format",
"seed",
+ "stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_p"
]
}
},
{
- "id": "google/gemini-2.5-flash-lite",
- "name": "Google: Gemini 2.5 Flash Lite",
+ "id": "google/gemini-2.5-pro",
+ "name": "Google: Gemini 2.5 Pro",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-07-22 18:04:36 +0200",
+ "created_at": "2025-06-17 16:12:24 +0200",
"context_window": 1048576,
- "max_output_tokens": 65535,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -14562,14 +15267,14 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.39999999999999997,
- "cached_input_per_million": 0.024999999999999998
+ "input_per_million": 1.25,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 0.31
}
}
},
"metadata": {
- "description": "Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, \"thinking\" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. ",
+ "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -14586,7 +15291,7 @@
},
"top_provider": {
"context_length": 1048576,
- "max_completion_tokens": 65535,
+ "max_completion_tokens": 65536,
"is_moderated": false
},
"per_request_limits": null,
@@ -14606,13 +15311,13 @@
}
},
{
- "id": "google/gemini-2.5-flash-lite-preview-06-17",
- "name": "Google: Gemini 2.5 Flash Lite Preview 06-17",
+ "id": "google/gemini-2.5-pro-preview",
+ "name": "Google: Gemini 2.5 Pro Preview 06-05",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-06-17 17:23:51 +0200",
+ "created_at": "2025-06-05 17:27:37 +0200",
"context_window": 1048576,
- "max_output_tokens": 65535,
+ "max_output_tokens": 65536,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -14633,14 +15338,14 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.39999999999999997,
- "cached_input_per_million": 0.024999999999999998
+ "input_per_million": 1.25,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 0.31
}
}
},
"metadata": {
- "description": "Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, \"thinking\" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. ",
+ "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.\n",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -14657,7 +15362,7 @@
},
"top_provider": {
"context_length": 1048576,
- "max_completion_tokens": 65535,
+ "max_completion_tokens": 65536,
"is_moderated": false
},
"per_request_limits": null,
@@ -14677,20 +15382,20 @@
}
},
{
- "id": "google/gemini-2.5-pro",
- "name": "Google: Gemini 2.5 Pro",
+ "id": "google/gemini-2.5-pro-preview-05-06",
+ "name": "Google: Gemini 2.5 Pro Preview 05-06",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-06-17 16:12:24 +0200",
+ "created_at": "2025-05-07 02:41:53 +0200",
"context_window": 1048576,
- "max_output_tokens": 65536,
+ "max_output_tokens": 65535,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "file",
- "image",
"text",
- "audio"
+ "image",
+ "file",
+ "audio"
],
"output": [
"text"
@@ -14715,9 +15420,9 @@
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "file",
- "image",
"text",
+ "image",
+ "file",
"audio"
],
"output_modalities": [
@@ -14728,7 +15433,7 @@
},
"top_provider": {
"context_length": 1048576,
- "max_completion_tokens": 65536,
+ "max_completion_tokens": 65535,
"is_moderated": false
},
"per_request_limits": null,
@@ -14748,20 +15453,18 @@
}
},
{
- "id": "google/gemini-2.5-pro-preview",
- "name": "Google: Gemini 2.5 Pro Preview 06-05",
+ "id": "google/gemini-flash-1.5-8b",
+ "name": "Google: Gemini 1.5 Flash 8B",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-06-05 17:27:37 +0200",
- "context_window": 1048576,
- "max_output_tokens": 65536,
+ "created_at": "2024-10-03 02:00:00 +0200",
+ "context_window": 1000000,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output": [
"text"
@@ -14775,21 +15478,19 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.31
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15,
+ "cached_input_per_million": 0.01
}
}
},
"metadata": {
- "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.\n",
+ "description": "Gemini Flash 1.5 8B is optimized for speed and efficiency, offering enhanced performance in small prompt tasks like chat, transcription, and translation. With reduced latency, it is highly effective for real-time and large-scale operations. This model focuses on cost-effective solutions while maintaining high-quality results.\n\n[Click here to learn more about this model](https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
- "file",
- "image",
"text",
- "audio"
+ "image"
],
"output_modalities": [
"text"
@@ -14798,15 +15499,15 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 65536,
+ "context_length": 1000000,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
"max_tokens",
- "reasoning",
+ "presence_penalty",
"response_format",
"seed",
"stop",
@@ -14819,20 +15520,17 @@
}
},
{
- "id": "google/gemini-2.5-pro-preview-05-06",
- "name": "Google: Gemini 2.5 Pro Preview 05-06",
+ "id": "google/gemma-2-27b-it",
+ "name": "Google: Gemma 2 27B",
"provider": "openrouter",
"family": "google",
- "created_at": "2025-05-07 02:41:53 +0200",
- "context_window": 1048576,
- "max_output_tokens": 65535,
+ "created_at": "2024-07-13 02:00:00 +0200",
+ "context_window": 8192,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image",
- "file",
- "audio"
+ "text"
],
"output": [
"text"
@@ -14840,68 +15538,59 @@
},
"capabilities": [
"streaming",
- "function_calling",
"structured_output"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 10.0,
- "cached_input_per_million": 0.31
+ "input_per_million": 0.65,
+ "output_per_million": 0.65
}
}
},
"metadata": {
- "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.",
+ "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image",
- "file",
- "audio"
+ "text"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1048576,
- "max_completion_tokens": 65535,
+ "context_length": 8192,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "include_reasoning",
+ "frequency_penalty",
"max_tokens",
- "reasoning",
+ "presence_penalty",
"response_format",
- "seed",
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
"top_p"
]
}
},
{
- "id": "google/gemini-flash-1.5",
- "name": "Google: Gemini 1.5 Flash ",
+ "id": "google/gemma-2-9b-it",
+ "name": "Google: Gemma 2 9B",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-05-14 02:00:00 +0200",
- "context_window": 1000000,
+ "created_at": "2024-06-28 02:00:00 +0200",
+ "context_window": 8192,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -14909,66 +15598,66 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.075,
- "output_per_million": 0.3,
- "cached_input_per_million": 0.01875
+ "input_per_million": 0.01,
+ "output_per_million": 0.02
}
}
},
"metadata": {
- "description": "Gemini 1.5 Flash is a foundation model that performs well at a variety of multimodal tasks such as visual understanding, classification, summarization, and creating content from image, audio and video. It's adept at processing visual and text inputs such as photographs, documents, infographics, and screenshots.\n\nGemini 1.5 Flash is designed for high-volume, high-frequency tasks where cost and latency matter. On most common tasks, Flash achieves comparable quality to other Gemini Pro models at a significantly reduced cost. Flash is well-suited for applications like chat assistants and on-demand content generation where speed and scale matter.\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n#multimodal",
+ "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1000000,
+ "context_length": 8192,
"max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemini-flash-1.5-8b",
- "name": "Google: Gemini 1.5 Flash 8B",
+ "id": "google/gemma-2-9b-it:free",
+ "name": "Google: Gemma 2 9B (free)",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-10-03 02:00:00 +0200",
- "context_window": 1000000,
+ "created_at": "2024-06-28 02:00:00 +0200",
+ "context_window": 8192,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text",
- "image"
+ "text"
],
"output": [
"text"
@@ -14976,60 +15665,52 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.0375,
- "output_per_million": 0.15,
- "cached_input_per_million": 0.01
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Gemini Flash 1.5 8B is optimized for speed and efficiency, offering enhanced performance in small prompt tasks like chat, transcription, and translation. With reduced latency, it is highly effective for real-time and large-scale operations. This model focuses on cost-effective solutions while maintaining high-quality results.\n\n[Click here to learn more about this model](https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).",
+ "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
"architecture": {
- "modality": "text+image->text",
+ "modality": "text->text",
"input_modalities": [
- "text",
- "image"
+ "text"
],
"output_modalities": [
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 1000000,
+ "context_length": 8192,
"max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "repetition_penalty",
"seed",
"stop",
- "structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemini-pro-1.5",
- "name": "Google: Gemini 1.5 Pro",
+ "id": "google/gemma-3-12b-it",
+ "name": "Google: Gemma 3 12B",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-04-09 02:00:00 +0200",
- "context_window": 2000000,
+ "created_at": "2025-03-13 22:50:25 +0100",
+ "context_window": 96000,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
@@ -15043,19 +15724,19 @@
},
"capabilities": [
"streaming",
- "function_calling",
- "structured_output"
+ "structured_output",
+ "predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 1.25,
- "output_per_million": 5.0
+ "input_per_million": 0.04,
+ "output_per_million": 0.14
}
}
},
"metadata": {
- "description": "Google's latest multimodal model, supports image and video[0] in text or chat prompts.\n\nOptimized for language tasks including:\n\n- Code generation\n- Text generation\n- Text editing\n- Problem solving\n- Recommendations\n- Information extraction\n- Data extraction or generation\n- AI agents\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n* [0]: Video input is not available through OpenRouter at this time.",
+ "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)",
"architecture": {
"modality": "text+image->text",
"input_modalities": [
@@ -15066,41 +15747,46 @@
"text"
],
"tokenizer": "Gemini",
- "instruct_type": null
+ "instruct_type": "gemma"
},
"top_provider": {
- "context_length": 2000000,
+ "context_length": 96000,
"max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "repetition_penalty",
"response_format",
"seed",
"stop",
"structured_outputs",
"temperature",
- "tool_choice",
- "tools",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-2-27b-it",
- "name": "Google: Gemma 2 27B",
+ "id": "google/gemma-3-12b-it:free",
+ "name": "Google: Gemma 3 12B (free)",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-07-13 02:00:00 +0200",
- "context_window": 8192,
- "max_output_tokens": null,
+ "created_at": "2025-03-13 22:50:25 +0100",
+ "context_window": 32768,
+ "max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -15108,22 +15794,16 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "predicted_outputs"
],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.65,
- "output_per_million": 0.65
- }
- }
- },
+ "pricing": {},
"metadata": {
- "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
+ "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
@@ -15132,35 +15812,41 @@
"instruct_type": "gemma"
},
"top_provider": {
- "context_length": 8192,
- "max_completion_tokens": null,
+ "context_length": 32768,
+ "max_completion_tokens": 8192,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
- "response_format",
+ "repetition_penalty",
+ "seed",
"stop",
- "structured_outputs",
"temperature",
+ "top_k",
+ "top_logprobs",
"top_p"
]
}
},
{
- "id": "google/gemma-2-9b-it",
- "name": "Google: Gemma 2 9B",
+ "id": "google/gemma-3-27b-it",
+ "name": "Google: Gemma 3 27B",
"provider": "openrouter",
"family": "google",
- "created_at": "2024-06-28 02:00:00 +0200",
- "context_window": 8192,
+ "created_at": "2025-03-12 06:12:39 +0100",
+ "context_window": 96000,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -15174,270 +15860,15 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.01,
- "output_per_million": 0.02
+ "input_per_million": 0.07,
+ "output_per_million": 0.26
}
}
},
"metadata": {
- "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
+ "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)",
"architecture": {
- "modality": "text->text",
- "input_modalities": [
- "text"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
- },
- "top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 8192,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
- "max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "response_format",
- "seed",
- "stop",
- "temperature",
- "top_k",
- "top_logprobs",
- "top_p"
- ]
- }
- },
- {
- "id": "google/gemma-2-9b-it:free",
- "name": "Google: Gemma 2 9B (free)",
- "provider": "openrouter",
- "family": "google",
- "created_at": "2024-06-28 02:00:00 +0200",
- "context_window": 8192,
- "max_output_tokens": 8192,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "text"
- ],
- "output": [
- "text"
- ]
- },
- "capabilities": [
- "streaming",
- "predicted_outputs"
- ],
- "pricing": {},
- "metadata": {
- "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).",
- "architecture": {
- "modality": "text->text",
- "input_modalities": [
- "text"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
- },
- "top_provider": {
- "context_length": 8192,
- "max_completion_tokens": 8192,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
- "max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "seed",
- "stop",
- "temperature",
- "top_k",
- "top_logprobs",
- "top_p"
- ]
- }
- },
- {
- "id": "google/gemma-3-12b-it",
- "name": "Google: Gemma 3 12B",
- "provider": "openrouter",
- "family": "google",
- "created_at": "2025-03-13 22:50:25 +0100",
- "context_window": 96000,
- "max_output_tokens": 8192,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "text",
- "image"
- ],
- "output": [
- "text"
- ]
- },
- "capabilities": [
- "streaming",
- "structured_output",
- "predicted_outputs"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.04,
- "output_per_million": 0.14
- }
- }
- },
- "metadata": {
- "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)",
- "architecture": {
- "modality": "text+image->text",
- "input_modalities": [
- "text",
- "image"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
- },
- "top_provider": {
- "context_length": 96000,
- "max_completion_tokens": 8192,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
- "max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "response_format",
- "seed",
- "stop",
- "temperature",
- "top_k",
- "top_logprobs",
- "top_p"
- ]
- }
- },
- {
- "id": "google/gemma-3-12b-it:free",
- "name": "Google: Gemma 3 12B (free)",
- "provider": "openrouter",
- "family": "google",
- "created_at": "2025-03-13 22:50:25 +0100",
- "context_window": 32768,
- "max_output_tokens": 8192,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "text",
- "image"
- ],
- "output": [
- "text"
- ]
- },
- "capabilities": [
- "streaming",
- "predicted_outputs"
- ],
- "pricing": {},
- "metadata": {
- "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)",
- "architecture": {
- "modality": "text+image->text",
- "input_modalities": [
- "text",
- "image"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Gemini",
- "instruct_type": "gemma"
- },
- "top_provider": {
- "context_length": 32768,
- "max_completion_tokens": 8192,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
- "max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
- "seed",
- "stop",
- "temperature",
- "top_k",
- "top_logprobs",
- "top_p"
- ]
- }
- },
- {
- "id": "google/gemma-3-27b-it",
- "name": "Google: Gemma 3 27B",
- "provider": "openrouter",
- "family": "google",
- "created_at": "2025-03-12 06:12:39 +0100",
- "context_window": 96000,
- "max_output_tokens": 8192,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "text",
- "image"
- ],
- "output": [
- "text"
- ]
- },
- "capabilities": [
- "streaming",
- "structured_output",
- "predicted_outputs"
- ],
- "pricing": {
- "text_tokens": {
- "standard": {
- "input_per_million": 0.07,
- "output_per_million": 0.26
- }
- }
- },
- "metadata": {
- "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)",
- "architecture": {
- "modality": "text+image->text",
+ "modality": "text+image->text",
"input_modalities": [
"text",
"image"
@@ -16662,62 +17093,8 @@
}
},
{
- "id": "meta-llama/llama-3.1-405b-instruct:free",
- "name": "Meta: Llama 3.1 405B Instruct (free)",
- "provider": "openrouter",
- "family": "meta-llama",
- "created_at": "2024-07-23 02:00:00 +0200",
- "context_window": 65536,
- "max_output_tokens": null,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [
- "text"
- ],
- "output": [
- "text"
- ]
- },
- "capabilities": [
- "streaming",
- "structured_output"
- ],
- "pricing": {},
- "metadata": {
- "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models including GPT-4o and Claude 3.5 Sonnet in evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).",
- "architecture": {
- "modality": "text->text",
- "input_modalities": [
- "text"
- ],
- "output_modalities": [
- "text"
- ],
- "tokenizer": "Llama3",
- "instruct_type": "llama3"
- },
- "top_provider": {
- "context_length": 65536,
- "max_completion_tokens": null,
- "is_moderated": false
- },
- "per_request_limits": null,
- "supported_parameters": [
- "frequency_penalty",
- "max_tokens",
- "presence_penalty",
- "response_format",
- "stop",
- "structured_outputs",
- "temperature",
- "top_k",
- "top_p"
- ]
- }
- },
- {
- "id": "meta-llama/llama-3.1-70b-instruct",
- "name": "Meta: Llama 3.1 70B Instruct",
+ "id": "meta-llama/llama-3.1-70b-instruct",
+ "name": "Meta: Llama 3.1 70B Instruct",
"provider": "openrouter",
"family": "meta-llama",
"created_at": "2024-07-23 02:00:00 +0200",
@@ -17206,8 +17583,8 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.012,
- "output_per_million": 0.036
+ "input_per_million": 0.04,
+ "output_per_million": 0.12
}
}
},
@@ -21088,9 +21465,7 @@
]
},
"capabilities": [
- "streaming",
- "function_calling",
- "predicted_outputs"
+ "streaming"
],
"pricing": {},
"metadata": {
@@ -21109,25 +21484,14 @@
"top_provider": {
"context_length": 32768,
"max_completion_tokens": null,
- "is_moderated": false
+ "is_moderated": true
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
- "logit_bias",
- "logprobs",
"max_tokens",
- "min_p",
- "presence_penalty",
- "repetition_penalty",
"seed",
"stop",
- "temperature",
- "tool_choice",
- "tools",
- "top_k",
- "top_logprobs",
- "top_p"
+ "temperature"
]
}
},
@@ -21565,6 +21929,71 @@
]
}
},
+ {
+ "id": "nousresearch/deephermes-3-llama-3-8b-preview",
+ "name": "Nous: DeepHermes 3 Llama 3 8B Preview",
+ "provider": "openrouter",
+ "family": "nousresearch",
+ "created_at": "2025-02-28 06:09:32 +0100",
+ "context_window": 131072,
+ "max_output_tokens": null,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "predicted_outputs"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.01,
+ "output_per_million": 0.049999999999999996
+ }
+ }
+ },
+ "metadata": {
+ "description": "DeepHermes 3 Preview is the latest version of our flagship Hermes series of LLMs by Nous Research, and one of the first models in the world to unify Reasoning (long chains of thought that improve answer accuracy) and normal LLM response modes into one model. We have also improved LLM annotation, judgement, and function calling.\n\nDeepHermes 3 Preview is one of the first LLM models to unify both \"intuitive\", traditional mode responses and long chain of thought reasoning responses into a single model, toggled by a system prompt.",
+ "architecture": {
+ "modality": "text->text",
+ "input_modalities": [
+ "text"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Other",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 131072,
+ "max_completion_tokens": null,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "frequency_penalty",
+ "logit_bias",
+ "logprobs",
+ "max_tokens",
+ "min_p",
+ "presence_penalty",
+ "repetition_penalty",
+ "seed",
+ "stop",
+ "temperature",
+ "top_k",
+ "top_logprobs",
+ "top_p"
+ ]
+ }
+ },
{
"id": "nousresearch/deephermes-3-llama-3-8b-preview:free",
"name": "Nous: DeepHermes 3 Llama 3 8B Preview (free)",
@@ -23956,9 +24385,9 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.625,
- "output_per_million": 5.0,
- "cached_input_per_million": 0.0625
+ "input_per_million": 1.25,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 0.125
}
}
},
@@ -24056,6 +24485,76 @@
]
}
},
+ {
+ "id": "openai/gpt-5-codex",
+ "name": "OpenAI: GPT-5 Codex",
+ "provider": "openrouter",
+ "family": "openai",
+ "created_at": "2025-09-23 18:03:23 +0200",
+ "context_window": 400000,
+ "max_output_tokens": 128000,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 1.25,
+ "output_per_million": 10.0,
+ "cached_input_per_million": 0.125
+ }
+ }
+ },
+ "metadata": {
+ "description": "GPT-5-Codex is a specialized version of GPT-5 optimized for software engineering and coding workflows. It is designed for both interactive development sessions and long, independent execution of complex engineering tasks. The model supports building projects from scratch, feature development, debugging, large-scale refactoring, and code review. Compared to GPT-5, Codex is more steerable, adheres closely to developer instructions, and produces cleaner, higher-quality code outputs. Reasoning effort can be adjusted with the `reasoning.effort` parameter. Read the [docs here](https://openrouter.ai/docs/use-cases/reasoning-tokens#reasoning-effort-level)\n\nCodex integrates into developer environments including the CLI, IDE extensions, GitHub, and cloud tasks. It adapts reasoning effort dynamically—providing fast responses for small tasks while sustaining extended multi-hour runs for large projects. The model is trained to perform structured code reviews, catching critical flaws by reasoning over dependencies and validating behavior against tests. It also supports multimodal inputs such as images or screenshots for UI development and integrates tool use for search, dependency installation, and environment setup. Codex is intended specifically for agentic coding applications.",
+ "architecture": {
+ "modality": "text+image->text",
+ "input_modalities": [
+ "text",
+ "image"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "GPT",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 400000,
+ "max_completion_tokens": 128000,
+ "is_moderated": true
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
+ "max_tokens",
+ "presence_penalty",
+ "reasoning",
+ "response_format",
+ "seed",
+ "stop",
+ "structured_outputs",
+ "tool_choice",
+ "tools",
+ "top_logprobs"
+ ]
+ }
+ },
{
"id": "openai/gpt-5-mini",
"name": "OpenAI: GPT-5 Mini",
@@ -26558,8 +27057,8 @@
"provider": "openrouter",
"family": "qwen",
"created_at": "2025-02-01 12:45:11 +0100",
- "context_window": 32768,
- "max_output_tokens": null,
+ "context_window": 131072,
+ "max_output_tokens": 2048,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -26590,21 +27089,17 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 2048,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
- "frequency_penalty",
"max_tokens",
"presence_penalty",
"response_format",
"seed",
- "stop",
- "structured_outputs",
"temperature",
- "top_k",
"top_p"
]
}
@@ -26616,7 +27111,7 @@
"family": "qwen",
"created_at": "2025-04-28 23:41:18 +0200",
"context_window": 40960,
- "max_output_tokens": 40960,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -26635,8 +27130,8 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.06,
- "output_per_million": 0.24
+ "input_per_million": 0.04,
+ "output_per_million": 0.14
}
}
},
@@ -26655,7 +27150,7 @@
},
"top_provider": {
"context_length": 40960,
- "max_completion_tokens": 40960,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -26822,7 +27317,7 @@
"family": "qwen",
"created_at": "2025-07-21 19:39:15 +0200",
"context_window": 262144,
- "max_output_tokens": 262144,
+ "max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
"input": [
@@ -26841,8 +27336,8 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.09999999999999999,
- "output_per_million": 0.09999999999999999
+ "input_per_million": 0.09,
+ "output_per_million": 0.6
}
}
},
@@ -26861,7 +27356,7 @@
},
"top_provider": {
"context_length": 262144,
- "max_completion_tokens": 262144,
+ "max_completion_tokens": null,
"is_moderated": false
},
"per_request_limits": null,
@@ -27768,7 +28263,7 @@
"name": "Qwen: Qwen3 Coder Plus",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-09-17 15:19:54 +0200",
+ "created_at": "2025-09-23 23:25:07 +0200",
"context_window": 128000,
"max_output_tokens": 65536,
"knowledge_cutoff": null,
@@ -27892,7 +28387,7 @@
"name": "Qwen: Qwen3 Max",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-09-05 14:49:27 +0200",
+ "created_at": "2025-09-23 23:26:48 +0200",
"context_window": 256000,
"max_output_tokens": 32768,
"knowledge_cutoff": null,
@@ -28094,17 +28589,18 @@
}
},
{
- "id": "qwen/qwq-32b",
- "name": "Qwen: QwQ 32B",
+ "id": "qwen/qwen3-vl-235b-a22b-instruct",
+ "name": "Qwen: Qwen3 VL 235B A22B Instruct",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2025-03-05 22:06:54 +0100",
- "context_window": 32768,
- "max_output_tokens": null,
+ "created_at": "2025-09-24 01:04:47 +0200",
+ "context_window": 131072,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -28119,27 +28615,28 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.15,
- "output_per_million": 0.39999999999999997
+ "input_per_million": 0.3,
+ "output_per_million": 1.5
}
}
},
"metadata": {
- "description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.",
+ "description": "Qwen3-VL-235B-A22B Instruct is an open-weight multimodal model that unifies strong text generation with visual understanding across images and video. The Instruct model targets general vision-language use (VQA, document parsing, chart/table extraction, multilingual OCR). The series emphasizes robust perception (recognition of diverse real-world and synthetic categories), spatial understanding (2D/3D grounding), and long-form visual comprehension, with competitive results on public multimodal benchmarks for both perception and reasoning.\n\nBeyond analysis, Qwen3-VL supports agentic interaction and tool use: it can follow complex instructions over multi-image, multi-turn dialogues; align text to video timelines for precise temporal queries; and operate GUI elements for automation tasks. The models also enable visual coding workflows—turning sketches or mockups into code and assisting with UI debugging—while maintaining strong text-only performance comparable to the flagship Qwen3 language models. This makes Qwen3-VL suitable for production scenarios spanning document AI, multilingual OCR, software/UI assistance, spatial/embodied tasks, and research on vision-language agents.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "qwq"
+ "tokenizer": "Qwen3",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 32768,
"is_moderated": false
},
"per_request_limits": null,
@@ -28147,7 +28644,6 @@
"frequency_penalty",
"include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@ -28161,23 +28657,23 @@
"tool_choice",
"tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwq-32b-preview",
- "name": "Qwen: QwQ 32B Preview",
+ "id": "qwen/qwen3-vl-235b-a22b-thinking",
+ "name": "Qwen: Qwen3 VL 235B A22B Thinking",
"provider": "openrouter",
"family": "qwen",
- "created_at": "2024-11-28 01:42:21 +0100",
- "context_window": 32768,
- "max_output_tokens": null,
+ "created_at": "2025-09-24 01:04:50 +0200",
+ "context_window": 131072,
+ "max_output_tokens": 32768,
"knowledge_cutoff": null,
"modalities": {
"input": [
- "text"
+ "text",
+ "image"
],
"output": [
"text"
@@ -28185,55 +28681,62 @@
},
"capabilities": [
"streaming",
+ "function_calling",
+ "structured_output",
"predicted_outputs"
],
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.19999999999999998,
- "output_per_million": 0.19999999999999998
+ "input_per_million": 0.3,
+ "output_per_million": 3.0
}
}
},
"metadata": {
- "description": "QwQ-32B-Preview is an experimental research model focused on AI reasoning capabilities developed by the Qwen Team. As a preview release, it demonstrates promising analytical abilities while having several important limitations:\n\n1. **Language Mixing and Code-Switching**: The model may mix languages or switch between them unexpectedly, affecting response clarity.\n2. **Recursive Reasoning Loops**: The model may enter circular reasoning patterns, leading to lengthy responses without a conclusive answer.\n3. **Safety and Ethical Considerations**: The model requires enhanced safety measures to ensure reliable and secure performance, and users should exercise caution when deploying it.\n4. **Performance and Benchmark Limitations**: The model excels in math and coding but has room for improvement in other areas, such as common sense reasoning and nuanced language understanding.\n\n",
+ "description": "Qwen3-VL-235B-A22B Thinking is a multimodal model that unifies strong text generation with visual understanding across images and video. The Thinking model is optimized for multimodal reasoning in STEM and math. The series emphasizes robust perception (recognition of diverse real-world and synthetic categories), spatial understanding (2D/3D grounding), and long-form visual comprehension, with competitive results on public multimodal benchmarks for both perception and reasoning.\n\nBeyond analysis, Qwen3-VL supports agentic interaction and tool use: it can follow complex instructions over multi-image, multi-turn dialogues; align text to video timelines for precise temporal queries; and operate GUI elements for automation tasks. The models also enable visual coding workflows, turning sketches or mockups into code and assisting with UI debugging, while maintaining strong text-only performance comparable to the flagship Qwen3 language models. This makes Qwen3-VL suitable for production scenarios spanning document AI, multilingual OCR, software/UI assistance, spatial/embodied tasks, and research on vision-language agents.",
"architecture": {
- "modality": "text->text",
+ "modality": "text+image->text",
"input_modalities": [
- "text"
+ "text",
+ "image"
],
"output_modalities": [
"text"
],
- "tokenizer": "Qwen",
- "instruct_type": "deepseek-r1"
+ "tokenizer": "Qwen3",
+ "instruct_type": null
},
"top_provider": {
- "context_length": 32768,
- "max_completion_tokens": null,
+ "context_length": 131072,
+ "max_completion_tokens": 32768,
"is_moderated": false
},
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
"logit_bias",
- "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
+ "reasoning",
"repetition_penalty",
+ "response_format",
"seed",
"stop",
+ "structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
- "top_logprobs",
"top_p"
]
}
},
{
- "id": "qwen/qwq-32b:free",
- "name": "Qwen: QwQ 32B (free)",
+ "id": "qwen/qwq-32b",
+ "name": "Qwen: QwQ 32B",
"provider": "openrouter",
"family": "qwen",
"created_at": "2025-03-05 22:06:54 +0100",
@@ -28250,9 +28753,18 @@
},
"capabilities": [
"streaming",
- "structured_output"
+ "function_calling",
+ "structured_output",
+ "predicted_outputs"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.15,
+ "output_per_million": 0.39999999999999997
+ }
+ }
+ },
"metadata": {
"description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.",
"architecture": {
@@ -28274,13 +28786,23 @@
"per_request_limits": null,
"supported_parameters": [
"frequency_penalty",
+ "include_reasoning",
+ "logit_bias",
+ "logprobs",
"max_tokens",
+ "min_p",
"presence_penalty",
+ "reasoning",
+ "repetition_penalty",
"response_format",
+ "seed",
"stop",
"structured_outputs",
"temperature",
+ "tool_choice",
+ "tools",
"top_k",
+ "top_logprobs",
"top_p"
]
}
@@ -28989,7 +29511,7 @@
"provider": "openrouter",
"family": "thedrummer",
"created_at": "2025-06-29 16:45:47 +0200",
- "context_window": 16384,
+ "context_window": 4096,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
@@ -29027,7 +29549,7 @@
"instruct_type": null
},
"top_provider": {
- "context_length": 16384,
+ "context_length": 4096,
"max_completion_tokens": null,
"is_moderated": false
},
@@ -30031,6 +30553,74 @@
]
}
},
+ {
+ "id": "x-ai/grok-4-fast",
+ "name": "xAI: Grok 4 Fast",
+ "provider": "openrouter",
+ "family": "x-ai",
+ "created_at": "2025-09-19 02:01:30 +0200",
+ "context_window": 2000000,
+ "max_output_tokens": 30000,
+ "knowledge_cutoff": null,
+ "modalities": {
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
+ },
+ "capabilities": [
+ "streaming",
+ "function_calling",
+ "structured_output"
+ ],
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.19999999999999998,
+ "output_per_million": 0.5,
+ "cached_input_per_million": 0.049999999999999996
+ }
+ }
+ },
+ "metadata": {
+ "description": "Grok 4 Fast is xAI's latest multimodal model with SOTA cost-efficiency and a 2M token context window. It comes in two flavors: non-reasoning and reasoning. Read more about the model on xAI's [news post](http://x.ai/news/grok-4-fast). Reasoning can be enabled using the `reasoning` `enabled` parameter in the API. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#controlling-reasoning-tokens)\n\nPrompts and completions may be used by xAI or OpenRouter to improve future models.",
+ "architecture": {
+ "modality": "text+image->text",
+ "input_modalities": [
+ "text",
+ "image"
+ ],
+ "output_modalities": [
+ "text"
+ ],
+ "tokenizer": "Grok",
+ "instruct_type": null
+ },
+ "top_provider": {
+ "context_length": 2000000,
+ "max_completion_tokens": 30000,
+ "is_moderated": false
+ },
+ "per_request_limits": null,
+ "supported_parameters": [
+ "include_reasoning",
+ "logprobs",
+ "max_tokens",
+ "reasoning",
+ "response_format",
+ "seed",
+ "structured_outputs",
+ "temperature",
+ "tool_choice",
+ "tools",
+ "top_logprobs",
+ "top_p"
+ ]
+ }
+ },
{
"id": "x-ai/grok-4-fast:free",
"name": "xAI: Grok 4 Fast (free)",
@@ -30241,8 +30831,8 @@
"pricing": {
"text_tokens": {
"standard": {
- "input_per_million": 0.41,
- "output_per_million": 1.6500000000000001
+ "input_per_million": 0.39999999999999997,
+ "output_per_million": 1.5999999999999999
}
}
},
@@ -30490,7 +31080,7 @@
"name": "Sonar",
"provider": "perplexity",
"family": "sonar",
- "created_at": "2025-09-21 16:12:52 +0200",
+ "created_at": "2025-09-26 05:44:27 +0200",
"context_window": 128000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
@@ -30522,7 +31112,7 @@
"name": "Sonar Deep Research",
"provider": "perplexity",
"family": "sonar_deep_research",
- "created_at": "2025-09-21 16:12:52 +0200",
+ "created_at": "2025-09-26 05:44:27 +0200",
"context_window": 128000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
@@ -30557,7 +31147,7 @@
"name": "Sonar Pro",
"provider": "perplexity",
"family": "sonar_pro",
- "created_at": "2025-09-21 16:12:52 +0200",
+ "created_at": "2025-09-26 05:44:27 +0200",
"context_window": 200000,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -30589,7 +31179,7 @@
"name": "Sonar Reasoning",
"provider": "perplexity",
"family": "sonar_reasoning",
- "created_at": "2025-09-21 16:12:52 +0200",
+ "created_at": "2025-09-26 05:44:27 +0200",
"context_window": 128000,
"max_output_tokens": 4096,
"knowledge_cutoff": null,
@@ -30621,7 +31211,7 @@
"name": "Sonar Reasoning Pro",
"provider": "perplexity",
"family": "sonar_reasoning_pro",
- "created_at": "2025-09-21 16:12:52 +0200",
+ "created_at": "2025-09-26 05:44:27 +0200",
"context_window": 128000,
"max_output_tokens": 8192,
"knowledge_cutoff": null,
@@ -30648,115 +31238,6 @@
},
"metadata": {}
},
- {
- "id": "chat-bison",
- "name": "chat-bison",
- "provider": "vertexai",
- "family": "palm",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [],
- "output": []
- },
- "capabilities": [
- "streaming"
- ],
- "pricing": {},
- "metadata": {
- "version_id": "002",
- "open_source_category": "PROPRIETARY",
- "launch_stage": "GA",
- "supported_actions": {
- "openGenie": {
- "references": {
- "us-central1": {
- "uri": "https://console.cloud.google.com/vertex-ai/generative/language/create/chat"
- }
- },
- "title": "Open Prompt Design"
- }
- },
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/chat-bison@002"
- }
- },
- {
- "id": "code-bison",
- "name": "code-bison",
- "provider": "vertexai",
- "family": "palm",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [],
- "output": []
- },
- "capabilities": [
- "streaming"
- ],
- "pricing": {},
- "metadata": {
- "version_id": "002",
- "open_source_category": null,
- "launch_stage": "GA",
- "supported_actions": null,
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/code-bison@002"
- }
- },
- {
- "id": "code-gecko",
- "name": "code-gecko",
- "provider": "vertexai",
- "family": "gemini",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [],
- "output": []
- },
- "capabilities": [
- "streaming"
- ],
- "pricing": {},
- "metadata": {
- "version_id": "002",
- "open_source_category": null,
- "launch_stage": "GA",
- "supported_actions": null,
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/code-gecko@002"
- }
- },
- {
- "id": "codechat-bison",
- "name": "codechat-bison",
- "provider": "vertexai",
- "family": "palm",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [],
- "output": []
- },
- "capabilities": [
- "streaming"
- ],
- "pricing": {},
- "metadata": {
- "version_id": "002",
- "open_source_category": null,
- "launch_stage": "GA",
- "supported_actions": null,
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/codechat-bison@002"
- }
- },
{
"id": "gemini-1.5-flash",
"name": "Gemini 1.5 Flash",
@@ -30777,6 +31258,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -30786,6 +31268,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -30813,6 +31299,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -30822,6 +31309,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -30849,6 +31340,7 @@
]
},
"capabilities": [
+ "batch",
"function_calling",
"structured_output"
],
@@ -30858,6 +31350,10 @@
"input_per_million": 0.075,
"cached_input_per_million": 0.01875,
"output_per_million": 0.3
+ },
+ "batch": {
+ "input_per_million": 0.0375,
+ "output_per_million": 0.15
}
}
},
@@ -31335,10 +31831,10 @@
}
},
{
- "id": "image-segmentation-001",
- "name": "image-segmentation-001",
+ "id": "text-embedding-004",
+ "name": "text-embedding-004",
"provider": "vertexai",
- "family": "gemini",
+ "family": "text-embedding",
"created_at": null,
"context_window": null,
"max_output_tokens": null,
@@ -31348,50 +31844,19 @@
"output": []
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling"
],
"pricing": {},
"metadata": {
- "version_id": "default",
- "open_source_category": null,
- "launch_stage": "PUBLIC_PREVIEW",
- "supported_actions": {
- "openNotebook": {
- "references": {
- "europe-west1": {
- "uri": "https://colab.research.google.com/github/GoogleCloudPlatform/generative-ai/blob/main/vision/getting-started/image_segmentation.ipynb"
- }
- },
- "title": "Open Notebook"
- },
- "requestAccess": {
- "references": {
- "europe-west1": {
- "uri": "https://docs.google.com/forms/d/e/1FAIpQLSdzIR1EeQGFcMsqd9nPip5e9ovDKSjfWRd58QVjo1zLpfdvEg/viewform?resourcekey=0-Pvqc66u-0Z1QmuzHq4wLKg"
- }
- }
- },
- "openNotebooks": {
- "notebooks": [
- {
- "references": {
- "europe-west1": {
- "uri": "https://colab.research.google.com/github/GoogleCloudPlatform/generative-ai/blob/main/vision/getting-started/image_segmentation.ipynb"
- }
- },
- "title": "Open Notebook"
- }
- ]
- }
- },
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/image-segmentation-001@default"
+ "source": "known_models"
}
},
{
- "id": "imagegeneration",
- "name": "imagegeneration",
+ "id": "text-embedding-005",
+ "name": "text-embedding-005",
"provider": "vertexai",
- "family": "gemini",
+ "family": "text-embedding",
"created_at": null,
"context_window": null,
"max_output_tokens": null,
@@ -31401,29 +31866,17 @@
"output": []
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling"
],
"pricing": {},
"metadata": {
- "version_id": "006",
- "open_source_category": "PROPRIETARY",
- "launch_stage": "PUBLIC_PREVIEW",
- "supported_actions": {
- "openGenerationAiStudio": {
- "references": {
- "europe-west1": {
- "uri": "https://cloud.google.com/console/vertex-ai/generative/vision"
- }
- },
- "title": "Open Vertex AI Studio"
- }
- },
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/imagegeneration@006"
+ "source": "known_models"
}
},
{
- "id": "imagen-4.0-fast-generate-001",
- "name": "imagen-4.0-fast-generate-001",
+ "id": "text-multilingual-embedding-002",
+ "name": "text-multilingual-embedding-002",
"provider": "vertexai",
"family": "gemini",
"created_at": null,
@@ -31435,326 +31888,361 @@
"output": []
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling"
],
"pricing": {},
"metadata": {
- "version_id": "default",
- "open_source_category": null,
- "launch_stage": "GA",
- "supported_actions": {
- "openGenerationAiStudio": {
- "references": {
- "europe-west1": {
- "uri": "https://console.cloud.google.com/vertex-ai/studio/media/generate"
- }
- },
- "title": "Open Vertex AI Studio"
- }
- },
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/imagen-4.0-fast-generate-001@default"
+ "source": "known_models"
}
},
{
- "id": "imagen-4.0-generate-001",
- "name": "imagen-4.0-generate-001",
- "provider": "vertexai",
- "family": "gemini",
+ "id": "grok-2-1212",
+ "name": "Grok-2 1212",
+ "provider": "xai",
+ "family": "grok_2",
"created_at": null,
- "context_window": null,
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [],
- "output": []
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
- "metadata": {
- "version_id": "default",
- "open_source_category": null,
- "launch_stage": "GA",
- "supported_actions": {
- "openGenerationAiStudio": {
- "references": {
- "europe-west1": {
- "uri": "https://console.cloud.google.com/vertex-ai/studio/media/generate"
- }
- },
- "title": "Open Vertex AI Studio"
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
}
- },
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/imagen-4.0-generate-001@default"
- }
- },
- {
- "id": "imagen-4.0-ultra-generate-001",
- "name": "imagen-4.0-ultra-generate-001",
- "provider": "vertexai",
- "family": "gemini",
- "created_at": null,
- "context_window": null,
- "max_output_tokens": null,
- "knowledge_cutoff": null,
- "modalities": {
- "input": [],
- "output": []
+ }
},
- "capabilities": [
- "streaming"
- ],
- "pricing": {},
"metadata": {
- "version_id": "default",
- "open_source_category": null,
- "launch_stage": "GA",
- "supported_actions": {
- "openGenerationAiStudio": {
- "references": {
- "europe-west1": {
- "uri": "https://console.cloud.google.com/vertex-ai/studio/media/generate"
- }
- },
- "title": "Open Vertex AI Studio"
- }
- },
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/imagen-4.0-ultra-generate-001@default"
+ "aliases": [
+ "grok-2",
+ "grok-2-latest"
+ ]
}
},
{
- "id": "imagetext",
- "name": "imagetext",
- "provider": "vertexai",
- "family": "gemini",
+ "id": "grok-2-image-1212",
+ "name": "Grok-2 Image 1212",
+ "provider": "xai",
+ "family": "grok_2",
"created_at": null,
- "context_window": null,
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [],
- "output": []
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "image",
+ "text"
+ ]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
- "metadata": {
- "version_id": "001",
- "open_source_category": "PROPRIETARY",
- "launch_stage": "GA",
- "supported_actions": {
- "openGenerationAiStudio": {
- "references": {
- "us-central1": {
- "uri": "https://cloud.google.com/console/vertex-ai/generative/vision"
- }
- },
- "title": "Open Vertex AI Studio"
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
}
- },
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/imagetext@001"
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-2-image",
+ "grok-2-image-latest"
+ ]
}
},
{
- "id": "multimodalembedding",
- "name": "multimodalembedding",
- "provider": "vertexai",
- "family": "gemini",
+ "id": "grok-2-vision-1212",
+ "name": "Grok-2 Vision 1212",
+ "provider": "xai",
+ "family": "grok_2_vision",
"created_at": null,
- "context_window": null,
+ "context_window": 32768,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [],
- "output": []
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
+ }
+ }
+ },
"metadata": {
- "version_id": "001",
- "open_source_category": "PROPRIETARY",
- "launch_stage": "GA",
- "supported_actions": null,
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/multimodalembedding@001"
+ "aliases": [
+ "grok-2-vision",
+ "grok-2-vision-latest"
+ ]
}
},
{
- "id": "text-bison",
- "name": "text-bison",
- "provider": "vertexai",
- "family": "palm",
+ "id": "grok-3",
+ "name": "Grok-3",
+ "provider": "xai",
+ "family": "grok_3",
"created_at": null,
- "context_window": null,
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [],
- "output": []
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
- "metadata": {
- "version_id": "002",
- "open_source_category": "PROPRIETARY",
- "launch_stage": "GA",
- "supported_actions": {
- "openGenie": {
- "references": {
- "us-central1": {
- "uri": "https://console.cloud.google.com/vertex-ai/generative/language/create/text"
- }
- },
- "title": "Open Prompt Design"
- },
- "openEvaluationPipeline": {
- "references": {
- "us-central1": {
- "uri": "https://console.cloud.google.com/vertex-ai/pipelines/vertex-ai-templates/evaluation-llm-text-generation-pipeline"
- }
- },
- "title": "Evaluate"
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
}
- },
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/text-bison@002"
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-3-latest",
+ "grok-3-beta",
+ "grok-3-fast",
+ "grok-3-fast-latest",
+ "grok-3-fast-beta"
+ ]
}
},
{
- "id": "text-embedding-004",
- "name": "text-embedding-004",
- "provider": "vertexai",
- "family": "text-embedding",
+ "id": "grok-3-mini",
+ "name": "Grok-3 Mini",
+ "provider": "xai",
+ "family": "grok_3_mini",
"created_at": null,
- "context_window": null,
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [],
- "output": []
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 0.3,
+ "output_per_million": 0.5,
+ "cached_input_per_million": 0.075
+ }
+ }
+ },
"metadata": {
- "source": "known_models"
+ "aliases": [
+ "grok-3-mini-latest",
+ "grok-3-mini-beta",
+ "grok-3-mini-fast",
+ "grok-3-mini-fast-latest",
+ "grok-3-mini-fast-beta"
+ ]
}
},
{
- "id": "text-embedding-005",
- "name": "text-embedding-005",
- "provider": "vertexai",
- "family": "text-embedding",
+ "id": "grok-4-0709",
+ "name": "Grok-4 0709",
+ "provider": "xai",
+ "family": "grok_4",
"created_at": null,
- "context_window": null,
+ "context_window": 256000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [],
- "output": []
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
+ }
+ }
+ },
"metadata": {
- "source": "known_models"
+ "aliases": [
+ "grok-4",
+ "grok-4-latest"
+ ]
}
},
{
- "id": "text-multilingual-embedding-002",
- "name": "text-multilingual-embedding-002",
- "provider": "vertexai",
- "family": "gemini",
+ "id": "grok-4-fast-non-reasoning",
+ "name": "Grok-4 Fast Non Reasoning",
+ "provider": "xai",
+ "family": "grok_4",
"created_at": null,
- "context_window": null,
+ "context_window": 256000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [],
- "output": []
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
},
"capabilities": [
"streaming",
- "function_calling"
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
+ }
+ }
+ },
"metadata": {
- "source": "known_models"
+ "aliases": [
+ "grok-4-fast-non-reasoning-latest"
+ ]
}
},
{
- "id": "text-unicorn",
- "name": "text-unicorn",
- "provider": "vertexai",
- "family": "gemini",
+ "id": "grok-4-fast-reasoning",
+ "name": "Grok-4 Fast Reasoning",
+ "provider": "xai",
+ "family": "grok_4",
"created_at": null,
- "context_window": null,
+ "context_window": 256000,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [],
- "output": []
+ "input": [
+ "text",
+ "image"
+ ],
+ "output": [
+ "text"
+ ]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
- "metadata": {
- "version_id": "001",
- "open_source_category": "PROPRIETARY",
- "launch_stage": "GA",
- "supported_actions": {
- "openGenie": {
- "references": {
- "europe-west1": {
- "uri": "https://console.cloud.google.com/vertex-ai/generative/language/create/text"
- }
- },
- "title": "Open in Vertex AI Studio"
- },
- "openEvaluationPipeline": {
- "references": {
- "europe-west1": {
- "uri": "https://console.cloud.google.com/vertex-ai/pipelines/vertex-ai-templates/evaluation-llm-text-generation-pipeline"
- }
- },
- "title": "Evaluate"
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 3.0,
+ "output_per_million": 15.0,
+ "cached_input_per_million": 0.75
}
- },
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/text-unicorn@001"
+ }
+ },
+ "metadata": {
+ "aliases": [
+ "grok-4-fast",
+ "grok-4-fast-reasoning-latest"
+ ]
}
},
{
- "id": "textembedding-gecko",
- "name": "textembedding-gecko",
- "provider": "vertexai",
- "family": "gemini",
+ "id": "grok-code-fast-1",
+ "name": "Grok-Code Fast 1",
+ "provider": "xai",
+ "family": "other",
"created_at": null,
- "context_window": null,
+ "context_window": 131072,
"max_output_tokens": null,
"knowledge_cutoff": null,
"modalities": {
- "input": [],
- "output": []
+ "input": [
+ "text"
+ ],
+ "output": [
+ "text"
+ ]
},
"capabilities": [
- "streaming"
+ "streaming",
+ "function_calling",
+ "structured_output"
],
- "pricing": {},
+ "pricing": {
+ "text_tokens": {
+ "standard": {
+ "input_per_million": 2.0,
+ "output_per_million": 10.0
+ }
+ }
+ },
"metadata": {
- "version_id": "003",
- "open_source_category": "PROPRIETARY",
- "launch_stage": "GA",
- "supported_actions": null,
- "publisher_model_template": "projects/{project}/locations/{location}/publishers/google/models/textembedding-gecko@003"
+ "aliases": [
+ "grok-code-fast",
+ "grok-code-fast-1-0825"
+ ]
}
}
]
\ No newline at end of file