Potential Breaking Change?

[bchau-calliope]

Bug Description

I have not encountered this error before regarding formatting and I could not find anything related on the repos, can someone investigate this?

File "/app/.venv/lib/python3.11/site-packages/livekit/agents/inference/llm.py", line 341, in _run\n raise APIStatusError(\nlivekit.agents._exceptions.APIStatusError: provider: openai model: gpt-4.1, message: POST "https://api.openai.com/v1/chat/completions\": 400 Bad Request {\n "message": "Missing required parameter: 'response_format.json_schema'.",\n "type": "invalid_request_error",\n "param": "response_format.json_schema",\n "code": "missing_required_parameter"\n }.

Expected Behavior

STT -> LLM -> TTS should work as expected.

Reproduction Steps

I think if you use openAI in the LLM node should encounter this.

Operating System

macOS Tahoe

Models Used

OpenAI GPT4.1 for LLM

Package Versions

all 1.3.10

Session/Room/Call IDs

No response

Proposed Solution

Additional Context

No response

Screenshots and Recordings

No response

[bchau-calliope]

Maybe its related to this? #4559

[chenghao-mou]

#4559 is not merged, so I don't think it is related. I can see you are using Inference, do you have any extra_kwargs applied? I think this only happens when you specify response_format but not provide the schema: extra_kwargs={"response_format": {"type": "json_schema"}}