I used to use the GPT-4o model, and it worked fine.
Yesterday, I tried the o1-mini model but got an invalid credential error. I already checked the deployment name and key.
I’m encountering the same issue with the Azure o1 model API not working, even though the 4o model API is functional. Is there an existing ticket for this problem, and any update on when it might be resolved?
Still experiencing problems with this! I definitely past the deployment name, e.g. ‘pleasework’, and if it’s o1-preview it doesn’t work, but if it’s gpt-4o it works
Hey all, it may be that our implementation here isn’t up to scratch to be able to be used with O1, but I’ll pass this on to the team and check what’s going on behind the scenes!
thank you because some of us have access to o1, it’s on Azure due to unrelated reasons, and we feel like we are one step away from being able to achieve amazing results. But that step has to be taken on your side
Cursor already implemented 100 models, but if you just add this advanced one that is so important to us, that if one of your devs is Christian, I’ll pray for the team, but if they swing another way I may sacrifice them a goat.
What is the ETA on this ?
This is happening because cursor is not specifying correct azure api version in backend .
Azure O1 requires latest API version .
Its 10 mins Fix , please include this in next update.
Thank you very much!
I’m not sure if this is the root cause, but I’d like to add my point in case it helps. Some parameters in the Azure o1 series API are different: max_completion_tokens has been added to support o1-preview and o1-mini models. The max_tokens parameter does not work with the o1 series models.
FYI Azure OpenAI Service API version lifecycle - Azure AI services | Microsoft Learn