[Fix] Support non-reasoning parameter for GPT-5 Chat model

docs/my-website/docs/providers/azure/azure.md

@@ -369,16 +369,20 @@ model_list:
 </Tabs>
 
 
-## GPT-5 Models
+## GPT-5 Reasoning Models
 
 | Property | Details |
 |-------|-------|
-| Description | Azure OpenAI GPT-5 models |
+| Description | Azure OpenAI GPT-5 reasoning models |
 | Provider Route on LiteLLM | `azure/gpt5_series/<custom-name>` or `azure/gpt-5-deployment-name` |
 
-LiteLLM supports using Azure GPT-5 models in one of the two ways:
+LiteLLM supports using Azure GPT-5 reasoning models in one of the two ways:
 1. Explicit Routing: `model = azure/gpt5_series/<deployment-name>`. In this scenario the model onboarded to litellm follows the format `model=azure/gpt5_series/<deployment-name>`.
-2. Inferred Routing (If the azure deployment name contains `gpt-5` in the name): `model = azure/gpt-5-mini`. In this scenario the model onboarded to litellm follows the format `model=azure/gpt-5-mini`.
+2. Inferred Routing (If the azure deployment name contains `gpt-5` but not `gpt-5-chat`): `model = azure/gpt-5-mini`. In this scenario the model onboarded to litellm follows the format `model=azure/gpt-5-mini`.
+
+:::note
+For GPT-5 chat models (e.g., `gpt-5-chat`), do not use the `gpt5_series/` prefix, as these are not reasoning models.
+:::
 
 #### Explicit Routing
 Use `azure/gpt5_series/<deployment-name>` for explicit GPT-5 model routing. 

litellm/__init__.py

@@ -1202,8 +1202,8 @@ def add_known_models():
 from .llms.openai.chat.gpt_transformation import (
     OpenAIGPTConfig,
 )
-from .llms.openai.chat.gpt_5_transformation import (
-    OpenAIGPT5Config,
+from .llms.openai.chat.gpt_5_reasoning_transformation import (
+    OpenAIGPT5ReasoningConfig,
 )
 from .llms.openai.transcriptions.whisper_transformation import (
     OpenAIWhisperAudioTranscriptionConfig,
@@ -1218,7 +1218,7 @@ def add_known_models():
 )
 
 openAIGPTAudioConfig = OpenAIGPTAudioConfig()
-openAIGPT5Config = OpenAIGPT5Config()
+openAIGPT5ReasoningConfig = OpenAIGPT5ReasoningConfig()
 
 from .llms.nvidia_nim.chat.transformation import NvidiaNimConfig
 from .llms.nvidia_nim.embed import NvidiaNimEmbeddingConfig
@@ -1256,7 +1256,7 @@ def add_known_models():
 from .llms.heroku.chat.transformation import HerokuChatConfig
 from .llms.cometapi.chat.transformation import CometAPIConfig
 from .llms.azure.chat.gpt_transformation import AzureOpenAIConfig
-from .llms.azure.chat.gpt_5_transformation import AzureOpenAIGPT5Config
+from .llms.azure.chat.gpt_5_reasoning_transformation import AzureOpenAIGPT5ReasoningConfig
 from .llms.azure.completion.transformation import AzureOpenAITextConfig
 from .llms.hosted_vllm.chat.transformation import HostedVLLMChatConfig
 from .llms.llamafile.chat.transformation import LlamafileChatConfig

litellm/litellm_core_utils/get_supported_openai_params.py

@@ -121,8 +121,8 @@ def get_supported_openai_params(  # noqa: PLR0915
             return litellm.AzureOpenAIO1Config().get_supported_openai_params(
                 model=model
             )
-        elif litellm.AzureOpenAIGPT5Config.is_model_gpt_5_model(model=model):
-            return litellm.AzureOpenAIGPT5Config().get_supported_openai_params(
+        elif litellm.AzureOpenAIGPT5ReasoningConfig.is_model_gpt_5_reasoning_model(model=model):
+            return litellm.AzureOpenAIGPT5ReasoningConfig().get_supported_openai_params(
                 model=model
             )
         else:

litellm/llms/azure/azure.py

@@ -230,8 +230,8 @@ def completion(  # noqa: PLR0915
                 )
 
                 data = {"model": None, "messages": messages, **optional_params}
-            elif litellm.AzureOpenAIGPT5Config.is_model_gpt_5_model(model=model):
-                data = litellm.AzureOpenAIGPT5Config().transform_request(
+            elif litellm.AzureOpenAIGPT5ReasoningConfig.is_model_gpt_5_reasoning_model(model=model):
+                data = litellm.AzureOpenAIGPT5ReasoningConfig().transform_request(
                     model=model,
                     messages=messages,
                     optional_params=optional_params,

litellm/llms/azure/chat/gpt_5_transformation.py → litellm/llms/azure/chat/gpt_5_reasoning_transformation.py

@@ -2,28 +2,28 @@
 
 from typing import List
 
-from litellm.llms.openai.chat.gpt_5_transformation import OpenAIGPT5Config
+from litellm.llms.openai.chat.gpt_5_reasoning_transformation import OpenAIGPT5ReasoningConfig
 from litellm.types.llms.openai import AllMessageValues
 
 from .gpt_transformation import AzureOpenAIConfig
 
 
-class AzureOpenAIGPT5Config(AzureOpenAIConfig, OpenAIGPT5Config):
-    """Azure specific handling for gpt-5 models."""
+class AzureOpenAIGPT5ReasoningConfig(AzureOpenAIConfig, OpenAIGPT5ReasoningConfig):
+    """Azure specific handling for gpt-5 reasoning models."""
 
     GPT5_SERIES_ROUTE = "gpt5_series/"
 
     @classmethod
-    def is_model_gpt_5_model(cls, model: str) -> bool:
+    def is_model_gpt_5_reasoning_model(cls, model: str) -> bool:
         """Check if the Azure model string refers to a gpt-5 variant.
 
         Accepts both explicit gpt-5 model names and the ``gpt5_series/`` prefix
         used for manual routing.
         """
-        return "gpt-5" in model or "gpt5_series" in model
+        return ("gpt-5" in model and "gpt-5-chat" not in model) or "gpt5_series" in model
 
     def get_supported_openai_params(self, model: str) -> List[str]:
-        return OpenAIGPT5Config.get_supported_openai_params(self, model=model)
+        return OpenAIGPT5ReasoningConfig.get_supported_openai_params(self, model=model)
 
     def map_openai_params(
         self,
@@ -33,7 +33,7 @@ def map_openai_params(
         drop_params: bool,
         api_version: str = "",
     ) -> dict:
-        return OpenAIGPT5Config.map_openai_params(
+        return OpenAIGPT5ReasoningConfig.map_openai_params(
             self,
             non_default_params=non_default_params,
             optional_params=optional_params,

litellm/llms/openai/chat/gpt_5_transformation.py → litellm/llms/openai/chat/gpt_5_reasoning_transformation.py

@@ -7,19 +7,19 @@
 from .gpt_transformation import OpenAIGPTConfig
 
 
-class OpenAIGPT5Config(OpenAIGPTConfig):
-    """Configuration for gpt-5 models including GPT-5-Codex variants.
+class OpenAIGPT5ReasoningConfig(OpenAIGPTConfig):
+    """Configuration for gpt-5 reasoning models including GPT-5-Codex variants.
 
-    Handles OpenAI API quirks for the gpt-5 series like:
+    Handles OpenAI API quirks for the gpt-5 reasoning series like:
 
     - Mapping ``max_tokens`` -> ``max_completion_tokens``.
     - Dropping unsupported ``temperature`` values when requested.
     - Support for GPT-5-Codex models optimized for code generation.
     """
 
     @classmethod
-    def is_model_gpt_5_model(cls, model: str) -> bool:
-        return "gpt-5" in model
+    def is_model_gpt_5_reasoning_model(cls, model: str) -> bool:
+        return "gpt-5" in model and "gpt-5-chat" not in model
 
     @classmethod
     def is_model_gpt_5_codex_model(cls, model: str) -> bool:

litellm/llms/openai/openai.py

@@ -51,7 +51,7 @@
 
 from ...types.llms.openai import *
 from ..base import BaseLLM
-from .chat.gpt_5_transformation import OpenAIGPT5Config
+from .chat.gpt_5_reasoning_transformation import OpenAIGPT5ReasoningConfig
 from .chat.o_series_transformation import OpenAIOSeriesConfig
 from .common_utils import (
     BaseOpenAILLM,
@@ -60,7 +60,7 @@
 )
 
 openaiOSeriesConfig = OpenAIOSeriesConfig()
-openAIGPT5Config = OpenAIGPT5Config()
+openAIGPT5ReasoningConfig = OpenAIGPT5ReasoningConfig()
 
 
 class MistralEmbeddingConfig:
@@ -189,8 +189,8 @@ def get_supported_openai_params(self, model: str) -> list:
         """
         if openaiOSeriesConfig.is_model_o_series_model(model=model):
             return openaiOSeriesConfig.get_supported_openai_params(model=model)
-        elif openAIGPT5Config.is_model_gpt_5_model(model=model):
-            return openAIGPT5Config.get_supported_openai_params(model=model)
+        elif openAIGPT5ReasoningConfig.is_model_gpt_5_reasoning_model(model=model):
+            return openAIGPT5ReasoningConfig.get_supported_openai_params(model=model)
         elif litellm.openAIGPTAudioConfig.is_model_gpt_audio_model(model=model):
             return litellm.openAIGPTAudioConfig.get_supported_openai_params(model=model)
         else:
@@ -225,8 +225,8 @@ def map_openai_params(
                 model=model,
                 drop_params=drop_params,
             )
-        elif openAIGPT5Config.is_model_gpt_5_model(model=model):
-            return openAIGPT5Config.map_openai_params(
+        elif openAIGPT5ReasoningConfig.is_model_gpt_5_reasoning_model(model=model):
+            return openAIGPT5ReasoningConfig.map_openai_params(
                 non_default_params=non_default_params,
                 optional_params=optional_params,
                 model=model,

litellm/model_prices_and_context_window_backup.json

@@ -1997,7 +1997,7 @@
         "supports_reasoning": true,
         "supports_response_schema": true,
         "supports_system_messages": true,
-        "supports_tool_choice": false,
+        "supports_tool_choice": true,
         "supports_vision": true
     },
     "azure/gpt-5-chat-latest": {
@@ -2029,7 +2029,7 @@
         "supports_reasoning": true,
         "supports_response_schema": true,
         "supports_system_messages": true,
-        "supports_tool_choice": false,
+        "supports_tool_choice": true,
         "supports_vision": true
     },
     "azure/gpt-5-codex": {

litellm/utils.py

@@ -4001,8 +4001,8 @@ def _check_valid_arg(supported_params: List[str]):
                     else False
                 ),
             )
-        elif litellm.AzureOpenAIGPT5Config.is_model_gpt_5_model(model=model):
-            optional_params = litellm.AzureOpenAIGPT5Config().map_openai_params(
+        elif litellm.AzureOpenAIGPT5ReasoningConfig.is_model_gpt_5_reasoning_model(model=model):
+            optional_params = litellm.AzureOpenAIGPT5ReasoningConfig().map_openai_params(
                 non_default_params=non_default_params,
                 optional_params=optional_params,
                 model=model,
@@ -6899,9 +6899,9 @@ def get_provider_chat_config(  # noqa: PLR0915
             return litellm.openaiOSeriesConfig
         elif (
             provider == LlmProviders.OPENAI
-            and litellm.OpenAIGPT5Config.is_model_gpt_5_model(model=model)
+            and litellm.OpenAIGPT5ReasoningConfig.is_model_gpt_5_reasoning_model(model=model)
         ):
-            return litellm.OpenAIGPT5Config()
+            return litellm.OpenAIGPT5ReasoningConfig()
         elif litellm.LlmProviders.DEEPSEEK == provider:
             return litellm.DeepSeekChatConfig()
         elif litellm.LlmProviders.GROQ == provider:
@@ -7012,8 +7012,8 @@ def get_provider_chat_config(  # noqa: PLR0915
         elif litellm.LlmProviders.AZURE == provider:
             if litellm.AzureOpenAIO1Config().is_o_series_model(model=model):
                 return litellm.AzureOpenAIO1Config()
-            if litellm.AzureOpenAIGPT5Config.is_model_gpt_5_model(model=model):
-                return litellm.AzureOpenAIGPT5Config()
+            if litellm.AzureOpenAIGPT5ReasoningConfig.is_model_gpt_5_reasoning_model(model=model):
+                return litellm.AzureOpenAIGPT5ReasoningConfig()
             return litellm.AzureOpenAIConfig()
         elif litellm.LlmProviders.AZURE_AI == provider:
             return litellm.AzureAIStudioConfig()

model_prices_and_context_window.json

@@ -1997,7 +1997,7 @@
         "supports_reasoning": true,
         "supports_response_schema": true,
         "supports_system_messages": true,
-        "supports_tool_choice": false,
+        "supports_tool_choice": true,
         "supports_vision": true
     },
     "azure/gpt-5-chat-latest": {
@@ -2029,7 +2029,7 @@
         "supports_reasoning": true,
         "supports_response_schema": true,
         "supports_system_messages": true,
-        "supports_tool_choice": false,
+        "supports_tool_choice": true,
         "supports_vision": true
     },
     "azure/gpt-5-codex": {

tests/test_litellm/llms/azure/chat/test_azure_gpt5_transformation.py → tests/test_litellm/llms/azure/chat/test_azure_gpt5_reasoning_transformation.py

@@ -1,22 +1,21 @@
 import pytest
 
 import litellm
-from litellm.llms.azure.chat.gpt_5_transformation import AzureOpenAIGPT5Config
+from litellm.llms.azure.chat.gpt_5_reasoning_transformation import AzureOpenAIGPT5ReasoningConfig
 
 
 @pytest.fixture()
-def config() -> AzureOpenAIGPT5Config:
-    return AzureOpenAIGPT5Config()
+def config() -> AzureOpenAIGPT5ReasoningConfig:
+    return AzureOpenAIGPT5ReasoningConfig()
 
 
-def test_azure_gpt5_supports_reasoning_effort(config: AzureOpenAIGPT5Config):
+def test_azure_gpt5_supports_reasoning_effort(config: AzureOpenAIGPT5ReasoningConfig):
     assert "reasoning_effort" in config.get_supported_openai_params(model="gpt-5")
     assert "reasoning_effort" in config.get_supported_openai_params(
         model="gpt5_series/my-deployment"
     )
 
-
-def test_azure_gpt5_maps_max_tokens(config: AzureOpenAIGPT5Config):
+def test_azure_gpt5_maps_max_tokens(config: AzureOpenAIGPT5ReasoningConfig):
     params = config.map_openai_params(
         non_default_params={"max_tokens": 5},
         optional_params={},
@@ -28,7 +27,7 @@ def test_azure_gpt5_maps_max_tokens(config: AzureOpenAIGPT5Config):
     assert "max_tokens" not in params
 
 
-def test_azure_gpt5_temperature_error(config: AzureOpenAIGPT5Config):
+def test_azure_gpt5_temperature_error(config: AzureOpenAIGPT5ReasoningConfig):
     with pytest.raises(litellm.utils.UnsupportedParamsError):
         config.map_openai_params(
             non_default_params={"temperature": 0.2},
@@ -38,8 +37,7 @@ def test_azure_gpt5_temperature_error(config: AzureOpenAIGPT5Config):
             api_version="2024-05-01-preview",
         )
 
-
-def test_azure_gpt5_series_transform_request(config: AzureOpenAIGPT5Config):
+def test_azure_gpt5_series_transform_request(config: AzureOpenAIGPT5ReasoningConfig):
     request = config.transform_request(
         model="gpt5_series/gpt-5",
         messages=[],
@@ -51,21 +49,21 @@ def test_azure_gpt5_series_transform_request(config: AzureOpenAIGPT5Config):
 
 
 # GPT-5-Codex specific tests for Azure
-def test_azure_gpt5_codex_model_detection(config: AzureOpenAIGPT5Config):
+def test_azure_gpt5_codex_model_detection(config: AzureOpenAIGPT5ReasoningConfig):
     """Test that Azure GPT-5-Codex models are correctly detected."""
-    assert config.is_model_gpt_5_model("gpt-5-codex")
-    assert config.is_model_gpt_5_model("gpt5_series/gpt-5-codex")
+    assert config.is_model_gpt_5_reasoning_model("gpt-5-codex")
+    assert config.is_model_gpt_5_reasoning_model("gpt5_series/gpt-5-codex")
 
 
-def test_azure_gpt5_codex_supports_reasoning_effort(config: AzureOpenAIGPT5Config):
+def test_azure_gpt5_codex_supports_reasoning_effort(config: AzureOpenAIGPT5ReasoningConfig):
     """Test that Azure GPT-5-Codex supports reasoning_effort parameter."""
     assert "reasoning_effort" in config.get_supported_openai_params(model="gpt-5-codex")
     assert "reasoning_effort" in config.get_supported_openai_params(
         model="gpt5_series/gpt-5-codex"
     )
 
 
-def test_azure_gpt5_codex_maps_max_tokens(config: AzureOpenAIGPT5Config):
+def test_azure_gpt5_codex_maps_max_tokens(config: AzureOpenAIGPT5ReasoningConfig):
     """Test that Azure GPT-5-Codex correctly maps max_tokens to max_completion_tokens."""
     params = config.map_openai_params(
         non_default_params={"max_tokens": 150},
@@ -78,7 +76,7 @@ def test_azure_gpt5_codex_maps_max_tokens(config: AzureOpenAIGPT5Config):
     assert "max_tokens" not in params
 
 
-def test_azure_gpt5_codex_temperature_error(config: AzureOpenAIGPT5Config):
+def test_azure_gpt5_codex_temperature_error(config: AzureOpenAIGPT5ReasoningConfig):
     """Test that Azure GPT-5-Codex raises error for unsupported temperature."""
     with pytest.raises(litellm.utils.UnsupportedParamsError):
         config.map_openai_params(
@@ -90,7 +88,7 @@ def test_azure_gpt5_codex_temperature_error(config: AzureOpenAIGPT5Config):
         )
 
 
-def test_azure_gpt5_codex_series_transform_request(config: AzureOpenAIGPT5Config):
+def test_azure_gpt5_codex_series_transform_request(config: AzureOpenAIGPT5ReasoningConfig):
     """Test that Azure GPT-5-Codex series routing works correctly."""
     request = config.transform_request(
         model="gpt5_series/gpt-5-codex",