Archit-Kumar-16 · Oct 1, 2026 · Sep 30, 2026 · Sep 30, 2026 · Oct 1, 2026
diff --git a/.github/scripts/autogen/types.ts b/.github/scripts/autogen/types.ts
@@ -157,6 +157,12 @@ export interface components {
             supportedModes?: components["schemas"]["Mode"][];
             /** @description Whether the model supports extended thinking / reasoning */
             thinking?: boolean;
+            /**
+             * @description When true, inbound Chat Completions stay on /chat/completions unless a
+             *     responses-only model or an existing trigger (header / reasoning / Cursor).
+             *     Omit on new models so Chat Completions is always served via Responses.
+             */
+            preferChatCompletions?: boolean;
         };
         ModelParam: {
             defaultValue?: (string | number | boolean) | null;

diff --git a/.github/test/model.cue b/.github/test/model.cue
@@ -341,6 +341,10 @@ package model
 	supportedModes?: [...#Mode]
 	// Whether the model supports extended thinking / reasoning
 	thinking?: bool
+	// When true, inbound Chat Completions stay on /chat/completions unless a
+	// responses-only model or an existing trigger (header / reasoning / Cursor).
+	// Omit on new models so Chat Completions is always served via Responses.
+	preferChatCompletions?: bool
 }
 
 #PricingTier: {

diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
@@ -147,6 +147,14 @@ removeParams: [<param-key>, ...]
 # Param keys that must always be provided by callers
 requiredParams: [<param-key>, ...]
 
+# Whether the model supports extended thinking / reasoning
+thinking: true
+
+# When true, inbound Chat Completions stay on /chat/completions unless a
+# responses-only model or an existing trigger (header / reasoning / Cursor).
+# Omit on new models so Chat Completions is always served via Responses.
+preferChatCompletions: true
+
 # Provider-specific escape-hatch configuration. Only set a key here when the
 # behaviour cannot be expressed through the generic fields above, and only on
 # models belonging to the provider that owns the key.

diff --git a/providers/aws-bedrock-mantle/google.gemma-4-26b-a4b.yaml b/providers/aws-bedrock-mantle/google.gemma-4-26b-a4b.yaml
@@ -30,6 +30,7 @@ modalities:
         - text
 mode: chat
 model: google.gemma-4-26b-a4b
+preferChatCompletions: true
 provisioning: serverless
 sources:
     - https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-4-26b-a4b.html

diff --git a/providers/aws-bedrock-mantle/google.gemma-4-31b.yaml b/providers/aws-bedrock-mantle/google.gemma-4-31b.yaml
@@ -27,6 +27,7 @@ modalities:
         - text
 mode: chat
 model: google.gemma-4-31b
+preferChatCompletions: true
 provisioning: serverless
 sources:
     - https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-4-31b.html

diff --git a/providers/aws-bedrock-mantle/google.gemma-4-e2b.yaml b/providers/aws-bedrock-mantle/google.gemma-4-e2b.yaml
@@ -28,6 +28,7 @@ modalities:
         - text
 mode: chat
 model: google.gemma-4-e2b
+preferChatCompletions: true
 provisioning: serverless
 sources:
     - https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-4-e2b.html

diff --git a/providers/aws-bedrock-mantle/openai.gpt-oss-120b.yaml b/providers/aws-bedrock-mantle/openai.gpt-oss-120b.yaml
@@ -69,6 +69,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 sources:
     - https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-120b.html

diff --git a/providers/aws-bedrock-mantle/openai.gpt-oss-20b.yaml b/providers/aws-bedrock-mantle/openai.gpt-oss-20b.yaml
@@ -61,6 +61,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 sources:
     - https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-20b.html

diff --git a/providers/aws-bedrock-mantle/xai.grok-4.3.yaml b/providers/aws-bedrock-mantle/xai.grok-4.3.yaml
@@ -51,6 +51,7 @@ params:
     - defaultValue: 131072
       key: max_completion_tokens
       type: number
+preferChatCompletions: true
 provisioning: serverless
 sources:
     - https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-xai-grok-4-3.html

diff --git a/providers/aws-bedrock-mantle/xai.grok-4.6.yaml b/providers/aws-bedrock-mantle/xai.grok-4.6.yaml
@@ -39,6 +39,7 @@ params:
     - defaultValue: 131072
       key: max_completion_tokens
       type: number
+preferChatCompletions: true
 provisioning: serverless
 sources:
     - https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-xai-grok-4-6.html

diff --git a/providers/aws-bedrock/openai.gpt-oss-120b-1:0.yaml b/providers/aws-bedrock/openai.gpt-oss-120b-1:0.yaml
@@ -84,5 +84,4 @@ sources:
 status: active
 supportedModes:
     - chat
-    - responses
 thinking: true
diff --git a/providers/aws-bedrock/openai.gpt-oss-20b-1:0.yaml b/providers/aws-bedrock/openai.gpt-oss-20b-1:0.yaml
@@ -85,5 +85,4 @@ sources:
 status: active
 supportedModes:
     - chat
-    - responses
 thinking: true
diff --git a/providers/azure-open-ai/computer-use-preview-2025-04-15.yaml b/providers/azure-open-ai/computer-use-preview-2025-04-15.yaml
@@ -27,6 +27,7 @@ params:
       key: max_tokens
       maxValue: 1024
       minValue: 1
+preferChatCompletions: true
 provisioning: serverless
 status: deprecated
 supportedModes:

diff --git a/providers/azure-open-ai/gpt-4.1-2025-04-14.yaml b/providers/azure-open-ai/gpt-4.1-2025-04-14.yaml
@@ -35,6 +35,7 @@ model: gpt-4.1-2025-04-14
 params:
     - key: max_tokens
       maxValue: 32768
+preferChatCompletions: true
 retirementDate: "2027-04-14"
 sources:
     - https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/models

diff --git a/providers/azure-open-ai/gpt-4.1-mini-2025-04-14.yaml b/providers/azure-open-ai/gpt-4.1-mini-2025-04-14.yaml
@@ -35,6 +35,7 @@ model: gpt-4.1-mini-2025-04-14
 params:
     - key: max_tokens
       maxValue: 32768
+preferChatCompletions: true
 provisioning: serverless
 retirementDate: "2027-04-14"
 sources:

diff --git a/providers/azure-open-ai/gpt-4.1-mini.yaml b/providers/azure-open-ai/gpt-4.1-mini.yaml
@@ -37,6 +37,7 @@ params:
       key: max_completion_tokens
       maxValue: 32768
       minValue: 1
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-4.1-nano-2025-04-14.yaml b/providers/azure-open-ai/gpt-4.1-nano-2025-04-14.yaml
@@ -37,6 +37,7 @@ params:
       key: max_tokens
       maxValue: 32768
       minValue: 1
+preferChatCompletions: true
 removeParams:
     - reasoning_effort
 retirementDate: "2027-04-14"

diff --git a/providers/azure-open-ai/gpt-4.1-nano.yaml b/providers/azure-open-ai/gpt-4.1-nano.yaml
@@ -40,6 +40,7 @@ params:
     - defaultValue: null
       key: response_format
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-4.1.yaml b/providers/azure-open-ai/gpt-4.1.yaml
@@ -46,6 +46,7 @@ params:
     - defaultValue: null
       key: response_format
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-4o-2024-05-13.yaml b/providers/azure-open-ai/gpt-4o-2024-05-13.yaml
@@ -23,6 +23,7 @@ params:
     - defaultValue: null
       key: response_format
       type: string
+preferChatCompletions: true
 supportedModes:
     - chat
     - responses
diff --git a/providers/azure-open-ai/gpt-4o-2024-08-06.yaml b/providers/azure-open-ai/gpt-4o-2024-08-06.yaml
@@ -33,6 +33,7 @@ params:
       key: max_tokens
       maxValue: 16384
       minValue: 1
+preferChatCompletions: true
 sources:
     - https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/models
 status: deprecated

diff --git a/providers/azure-open-ai/gpt-4o-2024-11-20.yaml b/providers/azure-open-ai/gpt-4o-2024-11-20.yaml
@@ -39,6 +39,7 @@ model: gpt-4o-2024-11-20
 params:
     - key: max_tokens
       maxValue: 16384
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - reasoning_effort

diff --git a/providers/azure-open-ai/gpt-4o-mini-2024-07-18.yaml b/providers/azure-open-ai/gpt-4o-mini-2024-07-18.yaml
@@ -38,6 +38,7 @@ model: gpt-4o-mini-2024-07-18
 params:
     - key: max_tokens
       maxValue: 16384
+preferChatCompletions: true
 provisioning: serverless
 retirementDate: "2027-04-14"
 status: deprecated

diff --git a/providers/azure-open-ai/gpt-4o-mini.yaml b/providers/azure-open-ai/gpt-4o-mini.yaml
@@ -40,6 +40,7 @@ params:
       key: max_tokens
       maxValue: 16384
       minValue: 1
+preferChatCompletions: true
 provisioning: serverless
 retirementDate: "2027-04-14"
 sources:

diff --git a/providers/azure-open-ai/gpt-4o.yaml b/providers/azure-open-ai/gpt-4o.yaml
@@ -35,6 +35,7 @@ model: gpt-4o
 params:
     - key: max_tokens
       maxValue: 16384
+preferChatCompletions: true
 provisioning: serverless
 retirementDate: "2026-10-01"
 status: deprecated

diff --git a/providers/azure-open-ai/gpt-5-2025-08-07.yaml b/providers/azure-open-ai/gpt-5-2025-08-07.yaml
@@ -49,6 +49,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5-chat-2025-08-07.yaml b/providers/azure-open-ai/gpt-5-chat-2025-08-07.yaml
@@ -31,6 +31,7 @@ params:
       key: max_completion_tokens
       maxValue: 16384
       minValue: 1
+preferChatCompletions: true
 removeParams:
     - max_tokens
 sources:

diff --git a/providers/azure-open-ai/gpt-5-chat-2025-10-03.yaml b/providers/azure-open-ai/gpt-5-chat-2025-10-03.yaml
@@ -28,6 +28,7 @@ model: gpt-5-chat-2025-10-03
 params:
     - key: max_completion_tokens
       maxValue: 16384
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5-chat-latest.yaml b/providers/azure-open-ai/gpt-5-chat-latest.yaml
@@ -29,6 +29,7 @@ params:
       key: max_completion_tokens
       maxValue: 16384
       minValue: 1
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5-chat.yaml b/providers/azure-open-ai/gpt-5-chat.yaml
@@ -30,6 +30,7 @@ params:
       key: max_completion_tokens
       maxValue: 16384
       minValue: 1
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5-mini-2025-08-07-lite.yaml b/providers/azure-open-ai/gpt-5-mini-2025-08-07-lite.yaml
@@ -43,6 +43,7 @@ params:
       key: max_completion_tokens
       maxValue: 128000
       minValue: 1
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5-mini-2025-08-07.yaml b/providers/azure-open-ai/gpt-5-mini-2025-08-07.yaml
@@ -57,6 +57,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5-mini.yaml b/providers/azure-open-ai/gpt-5-mini.yaml
@@ -56,6 +56,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - temperature

diff --git a/providers/azure-open-ai/gpt-5-nano-2025-08-07.yaml b/providers/azure-open-ai/gpt-5-nano-2025-08-07.yaml
@@ -45,6 +45,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5-nano.yaml b/providers/azure-open-ai/gpt-5-nano.yaml
@@ -58,6 +58,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - temperature

diff --git a/providers/azure-open-ai/gpt-5.1-2025-11-13.yaml b/providers/azure-open-ai/gpt-5.1-2025-11-13.yaml
@@ -59,6 +59,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - temperature

diff --git a/providers/azure-open-ai/gpt-5.1-chat-2025-11-13.yaml b/providers/azure-open-ai/gpt-5.1-chat-2025-11-13.yaml
@@ -38,6 +38,7 @@ params:
     - defaultValue: null
       key: reasoning_effort
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - temperature

diff --git a/providers/azure-open-ai/gpt-5.1-chat.yaml b/providers/azure-open-ai/gpt-5.1-chat.yaml
@@ -48,6 +48,7 @@ params:
           - high
           - xhigh
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - temperature

diff --git a/providers/azure-open-ai/gpt-5.1.yaml b/providers/azure-open-ai/gpt-5.1.yaml
@@ -59,6 +59,7 @@ params:
           - high
           - xhigh
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - temperature

diff --git a/providers/azure-open-ai/gpt-5.2-2025-12-11.yaml b/providers/azure-open-ai/gpt-5.2-2025-12-11.yaml
@@ -43,6 +43,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - temperature

diff --git a/providers/azure-open-ai/gpt-5.2-chat-2025-12-11.yaml b/providers/azure-open-ai/gpt-5.2-chat-2025-12-11.yaml
@@ -36,6 +36,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5.2-chat-2026-02-10.yaml b/providers/azure-open-ai/gpt-5.2-chat-2026-02-10.yaml
@@ -28,6 +28,7 @@ model: gpt-5.2-chat-2026-02-10
 params:
     - key: max_completion_tokens
       maxValue: 16384
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5.2-chat.yaml b/providers/azure-open-ai/gpt-5.2-chat.yaml
@@ -28,6 +28,7 @@ model: gpt-5.2-chat
 params:
     - key: max_completion_tokens
       maxValue: 16384
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5.2.yaml b/providers/azure-open-ai/gpt-5.2.yaml
@@ -47,6 +47,7 @@ params:
     - defaultValue: null
       key: response_format
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - temperature

diff --git a/providers/azure-open-ai/gpt-5.3-chat-2026-03-03.yaml b/providers/azure-open-ai/gpt-5.3-chat-2026-03-03.yaml
@@ -30,6 +30,7 @@ params:
       key: max_completion_tokens
       maxValue: 16384
       minValue: 1
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - max_tokens

diff --git a/providers/azure-open-ai/gpt-5.yaml b/providers/azure-open-ai/gpt-5.yaml
@@ -55,6 +55,7 @@ params:
           - medium
           - high
       type: string
+preferChatCompletions: true
 provisioning: serverless
 removeParams:
     - temperature

diff --git a/providers/azure-open-ai/gpt-chat-latest-2026-05-05.yaml b/providers/azure-open-ai/gpt-chat-latest-2026-05-05.yaml
@@ -23,6 +23,7 @@ modalities:
         - text
 mode: chat
 model: gpt-chat-latest-2026-05-05
+preferChatCompletions: true
 provisioning: serverless
 retirementDate: "2026-08-05"
 sources:

diff --git a/providers/azure-open-ai/gpt-chat-latest-2026-05-28.yaml b/providers/azure-open-ai/gpt-chat-latest-2026-05-28.yaml
@@ -25,6 +25,7 @@ modalities:
         - text
 mode: chat
 model: gpt-chat-latest-2026-05-28
+preferChatCompletions: true
 provisioning: serverless
 retirementDate: "2026-08-28"
 status: retired