[Bug]: LiteLLM Proxy doesn't drop temperature parameter for gpt-5 models when drop_params is enabled

[jannismain]

What happened?

Here are the relevant sections of my .litellm.yaml config:

model_list:
  - model_name: gpt-5
    litellm_params:
      model: "openai/gpt-5"
      api_key: ...
      drop_params: true

litellm_settings:
  drop_params: true

And this is the code I'm using:

import os, openai

client = openai.OpenAI(
    base_url="http://localhost:4000",
    api_key=os.getenv("LITELLM_PROXY_API_KEY"),
)

print(client.responses.create(input="Hi!", temperature=0.2, model="gpt-4o-mini").output_text)
print(client.responses.create(input="Hi!", temperature=0.2, model="gpt-5").output_text)

Expected

LiteLLM Proxy should respect drop_params: true and drop temperature parameter before issuing the call to OpenAI.

Actual

See log output below.

Note

With litellm Python SDK, drop_params=True does indeed drop the parameter for gpt-5 models. LiteLLM Proxy doesn't.

Relevant log output

Hello! How can I assist you today?
openai.BadRequestError: Error code: 400 - {'error': {'message': 'litellm.BadRequestError: OpenAIException - {\n  "error": {\n    "message": "Unsupported parameter: \'temperature\' is not supported with this model.",\n    "type": "invalid_request_error",\n    "param": "temperature",\n    "code": null\n  }\n}. Received Model Group=gpt-5\nAvailable Model Group Fallbacks=None', 'type': None, 'param': None, 'code': '400'}}

Are you a ML Ops Team?

No

What LiteLLM version are you on ?

v1.78.5

Twitter / LinkedIn details

No response