From a2837f1d740340c4f57b4fb7a88750f445aaa968 Mon Sep 17 00:00:00 2001
From: Lundin Matthews <lundinmatthews@me.com>
Date: Mon, 5 Jan 2026 18:45:38 -0500
Subject: [PATCH 1/2] Add LlamaGate provider with 26 open-source models

Adds LlamaGate (https://llamagate.dev) as a new provider with:
- 3 embedding models (Nomic, EmbeddingGemma, Qwen3 Embedding)
- 8 general purpose models (Llama 3.1/3.2, Mistral, Qwen, etc.)
- 6 code models (DeepSeek Coder, CodeLlama, Qwen Coder, etc.)
- 3 reasoning models (DeepSeek R1, OpenThinker)
- 5 vision models (Qwen VL, LLaVA, olmOCR, UI-TARS, Gemma 3)
- 1 tool calling model (FunctionGemma)

Provider details:
- API: OpenAI-compatible (@ai-sdk/openai-compatible)
- Base URL: https://api.llamagate.dev/v1
- Auth: Bearer token via LLAMAGATE_API_KEY
- Pricing: $0.02-$0.55 per 1M tokens
- All models are open-weights (Apache 2.0, MIT, Llama, Gemma)
---
 providers/llamagate/models/codegemma-7b.toml  | 21 ++++++++++++++++++
 providers/llamagate/models/codellama-7b.toml  | 21 ++++++++++++++++++
 .../llamagate/models/deepseek-coder-6.7b.toml | 21 ++++++++++++++++++
 .../llamagate/models/deepseek-r1-7b-qwen.toml | 21 ++++++++++++++++++
 .../llamagate/models/deepseek-r1-8b.toml      | 21 ++++++++++++++++++
 providers/llamagate/models/dolphin3-8b.toml   | 21 ++++++++++++++++++
 .../llamagate/models/embeddinggemma-300m.toml | 19 ++++++++++++++++
 providers/llamagate/models/functiongemma.toml | 21 ++++++++++++++++++
 providers/llamagate/models/gemma3-4b.toml     | 21 ++++++++++++++++++
 providers/llamagate/models/gpt-oss-20b.toml   | 21 ++++++++++++++++++
 providers/llamagate/models/granite4-3b.toml   | 21 ++++++++++++++++++
 providers/llamagate/models/llama-3.1-8b.toml  | 22 +++++++++++++++++++
 providers/llamagate/models/llama-3.2-3b.toml  | 22 +++++++++++++++++++
 providers/llamagate/models/llava-7b.toml      | 21 ++++++++++++++++++
 .../llamagate/models/mistral-7b-v0.3.toml     | 21 ++++++++++++++++++
 .../llamagate/models/nomic-embed-text.toml    | 19 ++++++++++++++++
 providers/llamagate/models/olmo-3-7b.toml     | 21 ++++++++++++++++++
 providers/llamagate/models/olmocr-7b.toml     | 21 ++++++++++++++++++
 .../llamagate/models/openthinker-7b.toml      | 21 ++++++++++++++++++
 .../llamagate/models/qwen2.5-coder-7b.toml    | 21 ++++++++++++++++++
 providers/llamagate/models/qwen3-8b.toml      | 21 ++++++++++++++++++
 .../llamagate/models/qwen3-embedding-8b.toml  | 19 ++++++++++++++++
 providers/llamagate/models/qwen3-vl-8b.toml   | 21 ++++++++++++++++++
 providers/llamagate/models/rnj-1-8b.toml      | 21 ++++++++++++++++++
 providers/llamagate/models/starcoder2-7b.toml | 21 ++++++++++++++++++
 providers/llamagate/models/ui-tars-7b.toml    | 21 ++++++++++++++++++
 providers/llamagate/provider.toml             |  5 +++++
 27 files changed, 547 insertions(+)
 create mode 100644 providers/llamagate/models/codegemma-7b.toml
 create mode 100644 providers/llamagate/models/codellama-7b.toml
 create mode 100644 providers/llamagate/models/deepseek-coder-6.7b.toml
 create mode 100644 providers/llamagate/models/deepseek-r1-7b-qwen.toml
 create mode 100644 providers/llamagate/models/deepseek-r1-8b.toml
 create mode 100644 providers/llamagate/models/dolphin3-8b.toml
 create mode 100644 providers/llamagate/models/embeddinggemma-300m.toml
 create mode 100644 providers/llamagate/models/functiongemma.toml
 create mode 100644 providers/llamagate/models/gemma3-4b.toml
 create mode 100644 providers/llamagate/models/gpt-oss-20b.toml
 create mode 100644 providers/llamagate/models/granite4-3b.toml
 create mode 100644 providers/llamagate/models/llama-3.1-8b.toml
 create mode 100644 providers/llamagate/models/llama-3.2-3b.toml
 create mode 100644 providers/llamagate/models/llava-7b.toml
 create mode 100644 providers/llamagate/models/mistral-7b-v0.3.toml
 create mode 100644 providers/llamagate/models/nomic-embed-text.toml
 create mode 100644 providers/llamagate/models/olmo-3-7b.toml
 create mode 100644 providers/llamagate/models/olmocr-7b.toml
 create mode 100644 providers/llamagate/models/openthinker-7b.toml
 create mode 100644 providers/llamagate/models/qwen2.5-coder-7b.toml
 create mode 100644 providers/llamagate/models/qwen3-8b.toml
 create mode 100644 providers/llamagate/models/qwen3-embedding-8b.toml
 create mode 100644 providers/llamagate/models/qwen3-vl-8b.toml
 create mode 100644 providers/llamagate/models/rnj-1-8b.toml
 create mode 100644 providers/llamagate/models/starcoder2-7b.toml
 create mode 100644 providers/llamagate/models/ui-tars-7b.toml
 create mode 100644 providers/llamagate/provider.toml

diff --git a/providers/llamagate/models/codegemma-7b.toml b/providers/llamagate/models/codegemma-7b.toml
new file mode 100644
index 00000000..0c3c70c5
--- /dev/null
+++ b/providers/llamagate/models/codegemma-7b.toml
@@ -0,0 +1,21 @@
+name = "CodeGemma 7B"
+release_date = "2024-04-09"
+last_updated = "2024-04-09"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.06
+output = 0.12
+
+[limit]
+context = 8_192
+output = 4_096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/codellama-7b.toml b/providers/llamagate/models/codellama-7b.toml
new file mode 100644
index 00000000..430fc1c9
--- /dev/null
+++ b/providers/llamagate/models/codellama-7b.toml
@@ -0,0 +1,21 @@
+name = "CodeLlama 7B"
+release_date = "2023-08-24"
+last_updated = "2023-08-24"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.06
+output = 0.12
+
+[limit]
+context = 16_384
+output = 4_096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/deepseek-coder-6.7b.toml b/providers/llamagate/models/deepseek-coder-6.7b.toml
new file mode 100644
index 00000000..eb980e8b
--- /dev/null
+++ b/providers/llamagate/models/deepseek-coder-6.7b.toml
@@ -0,0 +1,21 @@
+name = "DeepSeek Coder 6.7B"
+release_date = "2023-11-02"
+last_updated = "2023-11-02"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.06
+output = 0.12
+
+[limit]
+context = 16_384
+output = 4_096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/deepseek-r1-7b-qwen.toml b/providers/llamagate/models/deepseek-r1-7b-qwen.toml
new file mode 100644
index 00000000..9b43c7be
--- /dev/null
+++ b/providers/llamagate/models/deepseek-r1-7b-qwen.toml
@@ -0,0 +1,21 @@
+name = "DeepSeek R1 Distill Qwen 7B"
+release_date = "2025-01-20"
+last_updated = "2025-01-20"
+attachment = false
+reasoning = true
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.08
+output = 0.15
+
+[limit]
+context = 131_072
+output = 16_384
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/deepseek-r1-8b.toml b/providers/llamagate/models/deepseek-r1-8b.toml
new file mode 100644
index 00000000..0c2d3799
--- /dev/null
+++ b/providers/llamagate/models/deepseek-r1-8b.toml
@@ -0,0 +1,21 @@
+name = "DeepSeek R1 8B"
+release_date = "2025-01-20"
+last_updated = "2025-01-20"
+attachment = false
+reasoning = true
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.10
+output = 0.20
+
+[limit]
+context = 65_536
+output = 16_384
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/dolphin3-8b.toml b/providers/llamagate/models/dolphin3-8b.toml
new file mode 100644
index 00000000..d24caae5
--- /dev/null
+++ b/providers/llamagate/models/dolphin3-8b.toml
@@ -0,0 +1,21 @@
+name = "Dolphin 3 8B"
+release_date = "2024-07-23"
+last_updated = "2024-07-23"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.08
+output = 0.15
+
+[limit]
+context = 128_000
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/embeddinggemma-300m.toml b/providers/llamagate/models/embeddinggemma-300m.toml
new file mode 100644
index 00000000..54243e5f
--- /dev/null
+++ b/providers/llamagate/models/embeddinggemma-300m.toml
@@ -0,0 +1,19 @@
+name = "EmbeddingGemma 300M"
+release_date = "2024-06-27"
+last_updated = "2024-06-27"
+attachment = false
+reasoning = false
+temperature = false
+tool_call = false
+structured_output = false
+open_weights = true
+
+[cost]
+input = 0.02
+
+[limit]
+context = 2_048
+
+[modalities]
+input = ["text"]
+output = ["embedding"]
diff --git a/providers/llamagate/models/functiongemma.toml b/providers/llamagate/models/functiongemma.toml
new file mode 100644
index 00000000..2e2408a4
--- /dev/null
+++ b/providers/llamagate/models/functiongemma.toml
@@ -0,0 +1,21 @@
+name = "FunctionGemma"
+release_date = "2024-06-27"
+last_updated = "2024-06-27"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.02
+output = 0.04
+
+[limit]
+context = 8_192
+output = 2_048
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/gemma3-4b.toml b/providers/llamagate/models/gemma3-4b.toml
new file mode 100644
index 00000000..c9dc4981
--- /dev/null
+++ b/providers/llamagate/models/gemma3-4b.toml
@@ -0,0 +1,21 @@
+name = "Gemma 3 4B"
+release_date = "2025-03-12"
+last_updated = "2025-03-12"
+attachment = true
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.03
+output = 0.08
+
+[limit]
+context = 128_000
+output = 8_192
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/llamagate/models/gpt-oss-20b.toml b/providers/llamagate/models/gpt-oss-20b.toml
new file mode 100644
index 00000000..12cecafa
--- /dev/null
+++ b/providers/llamagate/models/gpt-oss-20b.toml
@@ -0,0 +1,21 @@
+name = "GPT-OSS 20B"
+release_date = "2024-12-01"
+last_updated = "2024-12-01"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.10
+output = 0.25
+
+[limit]
+context = 32_768
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/granite4-3b.toml b/providers/llamagate/models/granite4-3b.toml
new file mode 100644
index 00000000..0f5b38af
--- /dev/null
+++ b/providers/llamagate/models/granite4-3b.toml
@@ -0,0 +1,21 @@
+name = "Granite 4 3B"
+release_date = "2024-10-21"
+last_updated = "2024-10-21"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.04
+output = 0.08
+
+[limit]
+context = 128_000
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/llama-3.1-8b.toml b/providers/llamagate/models/llama-3.1-8b.toml
new file mode 100644
index 00000000..8f31ff74
--- /dev/null
+++ b/providers/llamagate/models/llama-3.1-8b.toml
@@ -0,0 +1,22 @@
+name = "Llama 3.1 8B Instruct"
+release_date = "2024-07-23"
+last_updated = "2024-07-23"
+knowledge = "2023-12"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.03
+output = 0.05
+
+[limit]
+context = 131_072
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/llama-3.2-3b.toml b/providers/llamagate/models/llama-3.2-3b.toml
new file mode 100644
index 00000000..848361a9
--- /dev/null
+++ b/providers/llamagate/models/llama-3.2-3b.toml
@@ -0,0 +1,22 @@
+name = "Llama 3.2 3B"
+release_date = "2024-09-25"
+last_updated = "2024-09-25"
+knowledge = "2023-12"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.04
+output = 0.08
+
+[limit]
+context = 131_072
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/llava-7b.toml b/providers/llamagate/models/llava-7b.toml
new file mode 100644
index 00000000..d6cc569e
--- /dev/null
+++ b/providers/llamagate/models/llava-7b.toml
@@ -0,0 +1,21 @@
+name = "LLaVA 1.5 7B"
+release_date = "2023-10-05"
+last_updated = "2023-10-05"
+attachment = true
+reasoning = false
+temperature = true
+tool_call = false
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.10
+output = 0.20
+
+[limit]
+context = 4_096
+output = 2_048
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/llamagate/models/mistral-7b-v0.3.toml b/providers/llamagate/models/mistral-7b-v0.3.toml
new file mode 100644
index 00000000..cc83baeb
--- /dev/null
+++ b/providers/llamagate/models/mistral-7b-v0.3.toml
@@ -0,0 +1,21 @@
+name = "Mistral 7B Instruct v0.3"
+release_date = "2024-05-22"
+last_updated = "2024-05-22"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.10
+output = 0.15
+
+[limit]
+context = 32_768
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/nomic-embed-text.toml b/providers/llamagate/models/nomic-embed-text.toml
new file mode 100644
index 00000000..f8091d22
--- /dev/null
+++ b/providers/llamagate/models/nomic-embed-text.toml
@@ -0,0 +1,19 @@
+name = "Nomic Embed Text"
+release_date = "2024-02-02"
+last_updated = "2024-02-02"
+attachment = false
+reasoning = false
+temperature = false
+tool_call = false
+structured_output = false
+open_weights = true
+
+[cost]
+input = 0.02
+
+[limit]
+context = 8_192
+
+[modalities]
+input = ["text"]
+output = ["embedding"]
diff --git a/providers/llamagate/models/olmo-3-7b.toml b/providers/llamagate/models/olmo-3-7b.toml
new file mode 100644
index 00000000..aef90485
--- /dev/null
+++ b/providers/llamagate/models/olmo-3-7b.toml
@@ -0,0 +1,21 @@
+name = "OLMo 3 7B"
+release_date = "2024-02-01"
+last_updated = "2024-02-01"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.10
+output = 0.20
+
+[limit]
+context = 8_192
+output = 4_096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/olmocr-7b.toml b/providers/llamagate/models/olmocr-7b.toml
new file mode 100644
index 00000000..0facb687
--- /dev/null
+++ b/providers/llamagate/models/olmocr-7b.toml
@@ -0,0 +1,21 @@
+name = "olmOCR 7B"
+release_date = "2024-02-25"
+last_updated = "2024-02-25"
+attachment = true
+reasoning = false
+temperature = true
+tool_call = false
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.08
+output = 0.15
+
+[limit]
+context = 8_192
+output = 4_096
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/llamagate/models/openthinker-7b.toml b/providers/llamagate/models/openthinker-7b.toml
new file mode 100644
index 00000000..fbeca8e7
--- /dev/null
+++ b/providers/llamagate/models/openthinker-7b.toml
@@ -0,0 +1,21 @@
+name = "OpenThinker 7B"
+release_date = "2024-11-01"
+last_updated = "2024-11-01"
+attachment = false
+reasoning = true
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.08
+output = 0.15
+
+[limit]
+context = 32_768
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/qwen2.5-coder-7b.toml b/providers/llamagate/models/qwen2.5-coder-7b.toml
new file mode 100644
index 00000000..d61ca498
--- /dev/null
+++ b/providers/llamagate/models/qwen2.5-coder-7b.toml
@@ -0,0 +1,21 @@
+name = "Qwen 2.5 Coder 7B"
+release_date = "2024-09-19"
+last_updated = "2024-09-19"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.06
+output = 0.12
+
+[limit]
+context = 32_768
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/qwen3-8b.toml b/providers/llamagate/models/qwen3-8b.toml
new file mode 100644
index 00000000..2ccb018c
--- /dev/null
+++ b/providers/llamagate/models/qwen3-8b.toml
@@ -0,0 +1,21 @@
+name = "Qwen 3 8B"
+release_date = "2024-09-19"
+last_updated = "2024-09-19"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.04
+output = 0.14
+
+[limit]
+context = 32_768
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/qwen3-embedding-8b.toml b/providers/llamagate/models/qwen3-embedding-8b.toml
new file mode 100644
index 00000000..b4dbf383
--- /dev/null
+++ b/providers/llamagate/models/qwen3-embedding-8b.toml
@@ -0,0 +1,19 @@
+name = "Qwen 3 Embedding 8B"
+release_date = "2025-01-01"
+last_updated = "2025-01-01"
+attachment = false
+reasoning = false
+temperature = false
+tool_call = false
+structured_output = false
+open_weights = true
+
+[cost]
+input = 0.02
+
+[limit]
+context = 40_960
+
+[modalities]
+input = ["text"]
+output = ["embedding"]
diff --git a/providers/llamagate/models/qwen3-vl-8b.toml b/providers/llamagate/models/qwen3-vl-8b.toml
new file mode 100644
index 00000000..67255e95
--- /dev/null
+++ b/providers/llamagate/models/qwen3-vl-8b.toml
@@ -0,0 +1,21 @@
+name = "Qwen 3 VL 8B"
+release_date = "2024-09-18"
+last_updated = "2024-09-18"
+attachment = true
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.15
+output = 0.55
+
+[limit]
+context = 32_768
+output = 8_192
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/llamagate/models/rnj-1-8b.toml b/providers/llamagate/models/rnj-1-8b.toml
new file mode 100644
index 00000000..aed96da4
--- /dev/null
+++ b/providers/llamagate/models/rnj-1-8b.toml
@@ -0,0 +1,21 @@
+name = "Rnj-1 8B"
+release_date = "2024-12-01"
+last_updated = "2024-12-01"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.10
+output = 0.20
+
+[limit]
+context = 32_000
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/starcoder2-7b.toml b/providers/llamagate/models/starcoder2-7b.toml
new file mode 100644
index 00000000..f52db2e6
--- /dev/null
+++ b/providers/llamagate/models/starcoder2-7b.toml
@@ -0,0 +1,21 @@
+name = "StarCoder2 7B"
+release_date = "2024-02-28"
+last_updated = "2024-02-28"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.06
+output = 0.12
+
+[limit]
+context = 16_384
+output = 4_096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/llamagate/models/ui-tars-7b.toml b/providers/llamagate/models/ui-tars-7b.toml
new file mode 100644
index 00000000..3fd863c5
--- /dev/null
+++ b/providers/llamagate/models/ui-tars-7b.toml
@@ -0,0 +1,21 @@
+name = "UI-TARS 1.5 7B"
+release_date = "2024-12-01"
+last_updated = "2024-12-01"
+attachment = true
+reasoning = false
+temperature = true
+tool_call = false
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.08
+output = 0.15
+
+[limit]
+context = 32_768
+output = 8_192
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/llamagate/provider.toml b/providers/llamagate/provider.toml
new file mode 100644
index 00000000..5d2b2789
--- /dev/null
+++ b/providers/llamagate/provider.toml
@@ -0,0 +1,5 @@
+name = "LlamaGate"
+npm = "@ai-sdk/openai-compatible"
+env = ["LLAMAGATE_API_KEY"]
+doc = "https://llamagate.dev/docs"
+api = "https://api.llamagate.dev/v1"

From 15076aa721b32ff60ceca621d8a48cce79489dbe Mon Sep 17 00:00:00 2001
From: Lundin Matthews <lundinmatthews@me.com>
Date: Tue, 6 Jan 2026 18:18:31 -0500
Subject: [PATCH 2/2] Fix embedding model schema validation
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

- Add output cost (0.00) for embedding models
- Add output limit (embedding dimension) for each model
- Change modalities.output from "embedding" to "text" (valid schema value)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
---
 providers/llamagate/models/embeddinggemma-300m.toml | 4 +++-
 providers/llamagate/models/nomic-embed-text.toml    | 4 +++-
 providers/llamagate/models/qwen3-embedding-8b.toml  | 4 +++-
 3 files changed, 9 insertions(+), 3 deletions(-)

diff --git a/providers/llamagate/models/embeddinggemma-300m.toml b/providers/llamagate/models/embeddinggemma-300m.toml
index 54243e5f..3fcf9dfe 100644
--- a/providers/llamagate/models/embeddinggemma-300m.toml
+++ b/providers/llamagate/models/embeddinggemma-300m.toml
@@ -10,10 +10,12 @@ open_weights = true
 
 [cost]
 input = 0.02
+output = 0.00
 
 [limit]
 context = 2_048
+output = 768
 
 [modalities]
 input = ["text"]
-output = ["embedding"]
+output = ["text"]
diff --git a/providers/llamagate/models/nomic-embed-text.toml b/providers/llamagate/models/nomic-embed-text.toml
index f8091d22..87a7046c 100644
--- a/providers/llamagate/models/nomic-embed-text.toml
+++ b/providers/llamagate/models/nomic-embed-text.toml
@@ -10,10 +10,12 @@ open_weights = true
 
 [cost]
 input = 0.02
+output = 0.00
 
 [limit]
 context = 8_192
+output = 768
 
 [modalities]
 input = ["text"]
-output = ["embedding"]
+output = ["text"]
diff --git a/providers/llamagate/models/qwen3-embedding-8b.toml b/providers/llamagate/models/qwen3-embedding-8b.toml
index b4dbf383..c1328221 100644
--- a/providers/llamagate/models/qwen3-embedding-8b.toml
+++ b/providers/llamagate/models/qwen3-embedding-8b.toml
@@ -10,10 +10,12 @@ open_weights = true
 
 [cost]
 input = 0.02
+output = 0.00
 
 [limit]
 context = 40_960
+output = 4_096
 
 [modalities]
 input = ["text"]
-output = ["embedding"]
+output = ["text"]