O1-preview API "request_timeout" error

API
1

Running o1-preview on a long series of request and hit this “request_timeout” error. Seems strange because the input is only ~2000 tokens and I believe the max_tokens argument has been depreciated for the o1 models. Anyone know why it might hit this error? If nothing else, the error message needs to be updated given the max_tokens argument is misleading :slight_smile:

"error": {                                                                                                                                                                                                                                                                                                 
        "message": "Timed out generating response. Please try again with a shorter prompt or with `max_tokens` set to a lower value.",
        "type": "internal_error",
        "param": null,
        "code": "request_timeout"
    }

Presuming it might have filled the output token limit with reasoning tokens and therefore not reached the output (?) but unsure (note: no max_completion_tokens parameter was set). Surely the API doesn’t fail if the reasoning tokens reach the limit?

If anyone has dealt with this error before would be much appriciated!