I have used the batch API for my application but have encountered with an unexpected error with status code 403 and error message “Project … does not have access to model gpt-4o-mini-2024-07-18-batch”. But when I check the list accessed models, I don’t see any model named “gpt-4o-mini-2024-07-18-batch”. How can I resolve the problem? Below are the details of the messages.
{
I have tried for 4 times since yesterday using batch api and i get the exact issue. I could not find the root cause as i ran the exact request 3 months ago and it worked. Can someone help me out?
Hi there! Not sure if you’re still experiencing this problem, but some others reported similar inconsistent behavior with Batch API and what worked for them was a combination of generating new API keys and making sure the model was enabled for usage on the Limits page.
If this still doesn’t work, you can create a new project to get past this bug. However if creating a new project is not reasonable (due to having to set everything up again), then I would recommend contacting OpenAI’s support team.
It must be due to a bug where the model gets tagged internally with this name “model-…-batch” but there’s another internal list of models somewhere that is not referencing it by the same name. This problem seems to only affect certain accounts / projects.