When using the new OpenAI responses API, the Gemini OpenAI Compatible Endpoint, as specified in the OpenAI compatibility section of the documentation, does not work.
It gives me an error that says error: 404 status code (no body).
I am unsure if the API is not intended to work, or if this is a bug.
This would be helpful in working on applications using the new Responses API. Thanks in advance!
Hi @EvanZhouDev , Welcome to the forum.
Responses API has not yet been implemented in the OpenAI compatible endpoint for Gemini. This is why you are encountering 404 error.
Thank you
Thank you for your quick response. Are there any plans to integrate the OpenAI Responses API in the near future?
Could you share a bit more about your use case and how you’d like to use it? We can raise this as a feature request with the product team.
Hi, just checking in to see if there are any plans to support the OpenAI Responses API? This would be extremely helpful, particularly for use cases like streaming+structured outputs.
This is really needed! This is the reason why we are not able to migrate to Gemini.
We are using OpenAI’s Agents SDK which is a powerful tool but we are not able to use Gemini with it since Responses API is not supported.
Agreed, I want to use gemini models but chat completions is the old api and has many limitations (mcps, file storage, attachments in prompt and other features) and also the performance is bad compared to responses api (speed, token management, structured outputs and severaal others).
I’m sticking with responses api limited to openai models until gemini models are compatible
I don’t think they care us! No answer to any related issue from Google side for months . I always get an answer from an OpenAI guy in a week in their forum.
No wonder why anybody is using Gemini API in production. They are months behind of OpenAI. AI development goes faster than what Google can support.