This model is only supported in v1/responses and not in v1/chat/completions

[VedantKhokher14]

When selecting o4-mini-deep-research in OpenAI models it shows the below error:

Something went wrong. Here's the specific error message we encountered: An error occurred while processing the request: 404 This model is only supported in v1/responses and not in v1/chat/completions. Troubleshooting URL: https://js.langchain.com/docs/troubleshooting/errors/MODEL_NOT_FOUND/

How to resolve it? Also is there some file I can make changes to?

[danny-avila]

Check the parameters panel, you can toggle on use of responses api

[martinounap]

Is there any way to make this easier for users? If I add these models to the list, it’s quite tedious to communicate it to everyone, and I’m pretty sure some will forget over time and come back with the same question.
I’ve also tried using modelSpecs, but that’s not an ideal solution in this case either.

[danny-avila]

you can add a model spec for this specific model and other, while also allowing the default provider > model options, including agents.

https://www.librechat.ai/docs/configuration/librechat_yaml/object_structure/model_specs#addedendpoints

[martinounap]

I’ll try to experiment with that. In my last test, it created a new model that didn’t appear under the OpenAI tab, which confused users. Also, when selecting a model under OpenAI, the Response API wasn’t enabled. However, when choosing the “new” model from the main list, it worked as expected.