GPT-4o-mini -- very slow and timeouts

I am finalizing an application that should go into production in two weeks.
GPT-4o-mini has become very slow lately and even goes to timeout. (GPT-4o is working but is far too expensive).

It used to be fast enough but since December, performances have dropped drastically.

Should I consider another provider than OpenAI?

1 Like

I’ll second this. Currently working on switching to a direct Azure integration because this is so unreliable.