com_error_invalid_request_error when trying to use o3 pro

[alexlyzhov]

What happened?

It says com_error_invalid_request_error w/o any further qualifications when I try to send a request to "o3-pro" openai model with default settings (I didn't change them).

People were saying in issues that this could happen for o3 earlier because organization is not verified, but my organization is verified (I see the badge in "Organization settings"), so that can't be the reason.

o3 model (not o3 pro) works fine.

Pressing "Reset model parameters" doesn't change any parameters or solve the issue.

Version Information

ghcr.io/danny-avila/librechat-dev latest c59eb01c308d 5 hours ago 1.79GB
ghcr.io/danny-avila/librechat-rag-api-dev-lite latest 2c984a6ef68f 28 hours ago 2.01GB

Steps to Reproduce

• update to latest version of Librechat (running using docker)
• add o3-pro into models list, reload the whole container
• pick it and try to send a request

What browsers are you seeing the problem on?

No response

Relevant log output

{"code":"document_not_found","httpStatus":404,"level":"debug","link":"https://docs.meilisearch.com/errors#document_not_found","message":"[MeiliMongooseModel.findOneAndUpdate] Convo not found in MeiliSearch and will index cbe6d3f9-3fb2-4bfb-9dfe-16708ba62032 Document `cbe6d3f9-3fb2-4bfb-9dfe-16708ba62032` not found.","name":"MeiliSearchApiError","stack":"MeiliSearchApiError: Document `cbe6d3f9-3fb2-4bfb-9dfe-16708ba62032` not found.\n    at /app/node_modules/meilisearch/dist/bundles/meilisearch.cjs.js:378:19\n    at Generator.next (<anonymous>)\n    at fulfilled (/app/node_modules/meilisearch/dist/bundles/meilisearch.cjs.js:298:58)\n    at process.processTicksAndRejections (node:internal/process/task_queues:95:5)","type":"invalid_request"}
{"code":"document_not_found","httpStatus":404,"level":"debug","link":"https://docs.meilisearch.com/errors#document_not_found","message":"[MeiliMongooseModel.findOneAndUpdate] Convo not found in MeiliSearch and will index 0dfc0cc9-d220-483f-ac51-478e2eefce09 Document `0dfc0cc9-d220-483f-ac51-478e2eefce09` not found.","name":"MeiliSearchApiError","stack":"MeiliSearchApiError: Document `0dfc0cc9-d220-483f-ac51-478e2eefce09` not found.\n    at /app/node_modules/meilisearch/dist/bundles/meilisearch.cjs.js:378:19\n    at Generator.next (<anonymous>)\n    at fulfilled (/app/node_modules/meilisearch/dist/bundles/meilisearch.cjs.js:298:58)\n    at process.processTicksAndRejections (node:internal/process/task_queues:95:5)","type":"invalid_request"}

Screenshots

No response

Code of Conduct

• I agree to follow this project's Code of Conduct

[danny-avila]

Nothing on our end that would cause this issue on latest:

Double-check the API key you are using.

Please also share debug logs:

Log locations:

• Docker: Project root directory ./logs
• npm: ./api/logs

There are two types of logs that can help diagnose the issue:

• debug logs (debug-YYYY-MM-DD.log)
• error logs (error-YYYY-MM-DD.log)

Error logs contain exact stack traces and are especially helpful, but both can provide valuable information.

[tinfever]

I might be seeing the same error when trying o3-pro. In the npm log I see "error: [AskController] Error handling request 404 This model is only supported in v1/responses and not in v1/chat/completions.". I see the open GitHub issue (#6364) regarding implementing the responses api but I don't totally understand the situation. Will o3-pro not work on LibreChat until this is implemented?

[danny-avila]

Will o3-pro not work on LibreChat until this is implemented?

Correct, OpenAI made it like this so that people move over to their new endpoint.

[sunsky89757]

The same issue for mine --- o3-pro(not o3)

[danny-avila]

See my comment above on how to share logs and which specific ones to share.

[enieuwy]

Same here

[bhavish-pj]

Hi @danny-avila,

I'm also encountering same issue with the o3-pro model that appears to be related to its API support.
My debug logs show this - "{"code":null,"message":"This model is only supported in v1/responses and not in v1/chat/completions.","param":"model","type":"invalid_request_error"}"
This aligns with the documentation I've found, which states that o3-pro exclusively supports the v1/responses API. You can find more details in their official documentation here: https://platform.openai.com/docs/models/o3-pro

Could you please guide me on the best way to resolve this?

TIA

[danny-avila]

o3-pro is not currently supported, please reread the discussion and the earlier comments.

[danny-avila]

o3-pro is now supported via Responses API toggle.