BaseModel to jsonschema for Structured Outputs

[federicoromeo]

Confirm this is a feature request for the Python library and not the underlying OpenAI API.

• This is a feature request for the Python library

Describe the feature or improvement you're requesting

Do you provide any helper function to generate the correct jsonschema starting from a pydantic BaseModel, to provide it to the response_format?

It's a big struggle not to have it. Imagine like in the Batch API in which i cannot provide the Pydantic instance

I tried with to_strict_json_schema but it doesn't work as expected, it puts keys like $defs and $refs.

Additional context

No response

[federicoromeo]

I found:
from openai.lib._parsing._completions import type_to_response_format_param

[HamzaYslmn]

from openai.resources.responses.responses import _type_to_text_format_param

[marcuswood-oai]

Thanks for the question and for sharing the helper. $defs and $ref are supported, so they do not need to be removed. For Batch Chat Completions, put type_to_response_format_param(YourModel) into each JSONL request's body.response_format; Batch uses the underlying endpoint's request parameters.

That import is a private-helper workaround. For a public-interface example, follow the Pydantic schema-export guidance in #2656 and use the Chat Completions json_schema wrapper. We’re closing this usage question with those steps; #2695 and #2993 remain open for a supported public helper. Please share a minimal schema and API error if it still fails.