diff --git a/modules/ai_foundry/README.md b/modules/ai_foundry/README.md index ab1144cf..9c98eab1 100644 --- a/modules/ai_foundry/README.md +++ b/modules/ai_foundry/README.md @@ -105,7 +105,7 @@ module "ai_foundry" { # Provide at least one model deployment model_deployments = [ - module.common_models.gpt_4_1, + module.common_models.gpt_5, module.common_models.text_embedding_3_large ] diff --git a/modules/common_models/README.md b/modules/common_models/README.md index 4bc7be54..c5fc412e 100644 --- a/modules/common_models/README.md +++ b/modules/common_models/README.md @@ -35,8 +35,8 @@ module "ai_foundry" { # ... other required inputs ... model_deployments = [ - module.common_models.gpt_4_1, - module.common_models.o4_mini, + module.common_models.gpt_5_2_chat, + module.common_models.gpt_5_nano, module.common_models.text_embedding_3_large ] } @@ -46,7 +46,7 @@ To override SKU per deployment: ```terraform model_deployments = [ - merge(module.common_models.gpt_4_1, { sku = { name = "GlobalStandard", capacity = 2 } }), + merge(module.common_models.gpt_5_2_chat, { sku = { name = "GlobalStandard", capacity = 2 } }), module.common_models.text_embedding_3_large ] ``` @@ -60,17 +60,24 @@ model_deployments = [ ## Outputs -| Name | Description | -|----------------------------|-------------------------------| -| gpt\_4\_1 | GPT-4.1 model | -| gpt\_4\_1\_mini | GPT-4.1-mini model | -| gpt\_4\_1\_nano | GPT-4.1-nano model | -| gpt\_4o | GPT-4o model | -| gpt\_4o\_audio\_preview | GPT-4o audio preview model | -| gpt\_4o\_mini | GPT-4o-mini model | -| gpt\_4o\_realtime\_preview | GPT-4o realtime preview model | -| gpt\_4o\_transcribe | GPT-4o transcribe model | -| o4\_mini | O4-mini model | -| text\_embedding\_3\_large | Text embedding 3 large model | -| text\_embedding\_3\_small | Text embedding 3 small model | +| Name | Description | +|------------------------------|--------------------------------------------------------------------------------| +| gpt\_4o\_mini\_transcribe | GPT-4o-mini-transcribe model - Improved transcription accuracy and robustness | +| gpt\_4o\_transcribe\_diarize | GPT-4o-transcribe-diarize model - Speech-to-text with speaker diarization | +| gpt\_5 | GPT-5 model - Reasoning model with advanced capabilities | +| gpt\_5\_1 | GPT-5.1 model - Enhanced reasoning with configurable reasoning\_effort | +| gpt\_5\_1\_chat | GPT-5.1-chat model - Built-in reasoning capabilities for chat | +| gpt\_5\_1\_codex | GPT-5.1-codex model - Optimized for Codex CLI and VS Code extension | +| gpt\_5\_1\_codex\_mini | GPT-5.1-codex-mini model - Lightweight codex model | +| gpt\_5\_2 | GPT-5.2 model - Reasoning model | +| gpt\_5\_2\_chat | GPT-5.2-chat model - Chat model with advanced capabilities | +| gpt\_5\_chat | GPT-5-chat model - Optimized for conversational AI with emotional intelligence | +| gpt\_5\_mini | GPT-5-mini model - Balanced performance and cost efficiency | +| gpt\_5\_nano | GPT-5-nano model - Lightweight model for high-throughput scenarios | +| gpt\_audio | GPT-audio model (GA) - Audio generation capabilities | +| gpt\_audio\_mini | GPT-audio-mini model - Lightweight audio generation | +| gpt\_realtime | GPT-realtime model (GA) - Real-time audio processing | +| gpt\_realtime\_mini | GPT-realtime-mini model - Lightweight real-time audio processing | +| text\_embedding\_3\_large | Text embedding 3 large model - Most capable embedding model | +| text\_embedding\_3\_small | Text embedding 3 small model | diff --git a/modules/common_models/outputs.tf b/modules/common_models/outputs.tf index c07100ba..4ff2dca3 100644 --- a/modules/common_models/outputs.tf +++ b/modules/common_models/outputs.tf @@ -2,58 +2,94 @@ # Copyright (c) Microsoft Corporation. Licensed under the MIT license. # --------------------------------------------------------------------- -output "gpt_4_1" { +output "gpt_5" { value = { format = "OpenAI" - name = "gpt-4.1" - version = "2025-04-14" + name = "gpt-5" + version = "2025-08-07" } - description = "GPT-4.1 model" + description = "GPT-5 model - Reasoning model with advanced capabilities" } -output "gpt_4_1_mini" { +output "gpt_5_mini" { value = { format = "OpenAI" - name = "gpt-4.1-mini" - version = "2025-04-14" + name = "gpt-5-mini" + version = "2025-08-07" } - description = "GPT-4.1-mini model" + description = "GPT-5-mini model - Balanced performance and cost efficiency" } -output "gpt_4_1_nano" { +output "gpt_5_nano" { value = { format = "OpenAI" - name = "gpt-4.1-nano" - version = "2025-04-14" + name = "gpt-5-nano" + version = "2025-08-07" } - description = "GPT-4.1-nano model" + description = "GPT-5-nano model - Lightweight model for high-throughput scenarios" } -output "gpt_4o" { +output "gpt_5_chat" { value = { format = "OpenAI" - name = "gpt-4o" - version = "2024-11-20" + name = "gpt-5-chat" + version = "2025-10-03" } - description = "GPT-4o model" + description = "GPT-5-chat model - Optimized for conversational AI with emotional intelligence" } -output "gpt_4o_mini" { +output "gpt_5_1" { value = { format = "OpenAI" - name = "gpt-4o-mini" - version = "2024-07-18" + name = "gpt-5.1" + version = "2025-11-13" } - description = "GPT-4o-mini model" + description = "GPT-5.1 model - Enhanced reasoning with configurable reasoning_effort" } -output "o4_mini" { +output "gpt_5_1_chat" { value = { format = "OpenAI" - name = "o4-mini" - version = "2025-04-16" + name = "gpt-5.1-chat" + version = "2025-11-13" } - description = "O4-mini model" + description = "GPT-5.1-chat model - Built-in reasoning capabilities for chat" +} + +output "gpt_5_1_codex" { + value = { + format = "OpenAI" + name = "gpt-5.1-codex" + version = "2025-11-13" + } + description = "GPT-5.1-codex model - Optimized for Codex CLI and VS Code extension" +} + +output "gpt_5_1_codex_mini" { + value = { + format = "OpenAI" + name = "gpt-5.1-codex-mini" + version = "2025-11-13" + } + description = "GPT-5.1-codex-mini model - Lightweight codex model" +} + +output "gpt_5_2" { + value = { + format = "OpenAI" + name = "gpt-5.2" + version = "2025-12-11" + } + description = "GPT-5.2 model - Reasoning model" +} + +output "gpt_5_2_chat" { + value = { + format = "OpenAI" + name = "gpt-5.2-chat" + version = "2025-12-11" + } + description = "GPT-5.2-chat model - Chat model with advanced capabilities" } output "text_embedding_3_small" { @@ -71,32 +107,59 @@ output "text_embedding_3_large" { name = "text-embedding-3-large" version = "1" } - description = "Text embedding 3 large model" + description = "Text embedding 3 large model - Most capable embedding model" +} + +output "gpt_realtime" { + value = { + format = "OpenAI" + name = "gpt-realtime" + version = "2025-08-28" + } + description = "GPT-realtime model (GA) - Real-time audio processing" +} + +output "gpt_realtime_mini" { + value = { + format = "OpenAI" + name = "gpt-realtime-mini" + version = "2025-10-06" + } + description = "GPT-realtime-mini model - Lightweight real-time audio processing" } -output "gpt_4o_realtime_preview" { +output "gpt_audio" { value = { format = "OpenAI" - name = "gpt-4o-realtime-preview" - version = "2025-06-03" + name = "gpt-audio" + version = "2025-08-28" } - description = "GPT-4o realtime preview model" + description = "GPT-audio model (GA) - Audio generation capabilities" } -output "gpt_4o_audio_preview" { +output "gpt_audio_mini" { value = { format = "OpenAI" - name = "gpt-4o-audio-preview" - version = "2024-12-17" + name = "gpt-audio-mini" + version = "2025-10-06" } - description = "GPT-4o audio preview model" + description = "GPT-audio-mini model - Lightweight audio generation" } -output "gpt_4o_transcribe" { +output "gpt_4o_mini_transcribe" { value = { format = "OpenAI" - name = "gpt-4o-transcribe" - version = "2025-03-20" + name = "gpt-4o-mini-transcribe" + version = "2025-12-15" + } + description = "GPT-4o-mini-transcribe model - Improved transcription accuracy and robustness" +} + +output "gpt_4o_transcribe_diarize" { + value = { + format = "OpenAI" + name = "gpt-4o-transcribe-diarize" + version = "1" } - description = "GPT-4o transcribe model" + description = "GPT-4o-transcribe-diarize model - Speech-to-text with speaker diarization" } diff --git a/reference_architectures/foundry_basic/main.tf b/reference_architectures/foundry_basic/main.tf index b4219ee5..2ee09a8b 100644 --- a/reference_architectures/foundry_basic/main.tf +++ b/reference_architectures/foundry_basic/main.tf @@ -58,8 +58,8 @@ module "ai_foundry" { # Model deployments to make available within Foundry # Add/remove models as needed for your workload requirements model_deployments = [ - module.common_models.gpt_4_1, - module.common_models.o4_mini, + module.common_models.gpt_5_2_chat, + module.common_models.gpt_5_nano, module.common_models.text_embedding_3_large ] diff --git a/reference_architectures/foundry_basic_private/main.tf b/reference_architectures/foundry_basic_private/main.tf index d4743d18..2ce5e4a8 100644 --- a/reference_architectures/foundry_basic_private/main.tf +++ b/reference_architectures/foundry_basic_private/main.tf @@ -45,8 +45,8 @@ module "ai_foundry" { name = module.naming.cognitive_account.name_unique model_deployments = [ - module.common_models.gpt_4_1, - module.common_models.o4_mini, + module.common_models.gpt_5_2_chat, + module.common_models.gpt_5_nano, module.common_models.text_embedding_3_large ] diff --git a/reference_architectures/foundry_standard/main.tf b/reference_architectures/foundry_standard/main.tf index 9f4fb964..57dee22d 100644 --- a/reference_architectures/foundry_standard/main.tf +++ b/reference_architectures/foundry_standard/main.tf @@ -59,8 +59,8 @@ module "ai_foundry" { # Model deployments to make available within Foundry # Add/remove models as needed for your workload requirements model_deployments = [ - module.common_models.gpt_4_1, - module.common_models.o4_mini, + module.common_models.gpt_5_2_chat, + module.common_models.gpt_5_nano, module.common_models.text_embedding_3_large ] diff --git a/reference_architectures/foundry_standard_private/main.tf b/reference_architectures/foundry_standard_private/main.tf index ab164f39..c7e27413 100644 --- a/reference_architectures/foundry_standard_private/main.tf +++ b/reference_architectures/foundry_standard_private/main.tf @@ -57,8 +57,8 @@ module "ai_foundry" { name = module.naming.cognitive_account.name_unique model_deployments = [ - module.common_models.gpt_4_1, - module.common_models.o4_mini, + module.common_models.gpt_5_2_chat, + module.common_models.gpt_5_nano, module.common_models.text_embedding_3_large ]