planning use_responses_api for AzureChatOpenAI and reasoning summary support?

[weipienlee]

Hi @baskaryan, it seems that ChatOpenAI already has the params below. Any idea when the Azure variant will be next? I would like to test whether or organization requires verification.

reasoning = {
    "effort": "medium",  # 'low', 'medium', or 'high'
    "summary": "auto",  # 'detailed', 'auto', or None
}

llm = AzureChatOpenAI(
    ...
    use_responses_api=True,
    model_kwargs={"reasoning": reasoning},
)

[ccurme]

Hi @weipienlee, usage through AzureChatOpenAI should be supported (though you need to set the API version > 2025-03-01-preview). Let me know if you're running into any specific issues.