[Feature]: Support 'Thinking' / 'reasoning' for hostes_vllm models

[puwang01]

Check for existing issues

• I have searched the existing issues and checked that my issue is not a duplicate.

The Feature

Enable the "thinking" support for the hosted_vllm/GLM-4.6-* and hosted_vllm/GLM-4.7-*.

Tried to use the "reasoning_effort" parameter in the endpoint /v1/chat/completions, works as expected.
But when using the endpoint /v1/messages, I got the error like below:

{'error': {'message': "litellm.UnsupportedParamsError: hosted_vllm does not support parameters: ['thinking'], for model=GLM-4.6-FP8. To drop these, set `litellm.drop_params=True` or for proxy:\n\n`litellm_settings:\n drop_params: true`\n. \n If you want to use these params dynamically send allowed_openai_params=['thinking'] in your request.. Received Model Group=glm-4.6\nAvailable Model Group Fallbacks=None", 'type': 'None', 'param': None, 'code': '400'}, 'status_code': 400, 'details': '{"error":{"message":"litellm.UnsupportedParamsError: hosted_vllm does not support parameters: [\'thinking\'], for model=GLM-4.6-FP8. To drop these, set `litellm.drop_params=True` or for proxy:\\n\\n`litellm_settings:\\n drop_params: true`\\n. \\n If you want to use these params dynamically send allowed_openai_params=[\'thinking\'] in your request.. Received Model Group=glm-4.6\\nAvailable Model Group Fallbacks=None","type":"None","param":null,"code":"400"}}'}

Motivation, pitch

I am using claude code cli to connect with my on premises deployed glm-4.6 model.

What part of LiteLLM is this about?

Proxy

LiteLLM is hiring a founding backend engineer, are you interested in joining us and shipping to all our users?

No

Twitter / LinkedIn details

No response