fix(openai): Fix Azure OpenAI Responses API model field issue (#32649)

libs/partners/openai/langchain_openai/chat_models/azure.py

@@ -754,6 +754,26 @@ class AzureChatOpenAI(BaseChatOpenAI):
 
         return chat_result
 
+    def _get_request_payload(
+        self,
+        input_: LanguageModelInput,
+        *,
+        stop: Optional[list[str]] = None,
+        **kwargs: Any,
+    ) -> dict:
+        """Get the request payload, using deployment name for Azure Responses API."""
+        payload = super()._get_request_payload(input_, stop=stop, **kwargs)
+
+        # For Azure Responses API, use deployment name instead of model name
+        if (
+            self._use_responses_api(payload)
+            and not payload.get("model")
+            and self.deployment_name
+        ):
+            payload["model"] = self.deployment_name
+
+        return payload
+
     def _stream(self, *args: Any, **kwargs: Any) -> Iterator[ChatGenerationChunk]:
         """Route to Chat Completions or Responses API."""
         if self._use_responses_api({**kwargs, **self.model_kwargs}):

libs/partners/openai/langchain_openai/chat_models/base.py

@@ -3541,7 +3541,7 @@ def _construct_responses_api_payload(
         payload["reasoning"] = {"effort": payload.pop("reasoning_effort")}
 
     # Remove temperature parameter for models that don't support it in responses API
-    model = payload.get("model", "")
+    model = payload.get("model") or ""
     if model.startswith("gpt-5") and "chat" not in model:  # gpt-5-chat supports
         payload.pop("temperature", None)