O3-mini-high performance degradation (feb 5th) [possibily requires oai staff attention]

so, just a few minutes ago o3-mini-high started struggling… output became shorter, its slower… I noticed there is nothing over at status.openai.com so I though it would be a good idea to make a post in case it starts affecting more users.

not sure if its because I’ve been using it for a while now, but I’m on a pro plan.

It’s now outputting maybe 1/5 of what is usually capable of and making way more mistakes then usual

oh, and this is on chatgpt, no idea about the api


image for reference of the status page

and yeah, its pretty bad, I don’t know if there is a limit on o3-mini-high for the pro tier… but it feels like o1-mini maybe… its thinking way less and the outputs are much, much shorter, with much less quality then what I’ve been seeing for the last few hours of use… it started on a chat, I tried a new chat and the same issue so far

the reasoning is also going a bit crazy, its started making random calculations out of no where… something is going on and I don’t know what it is

@vb @Foxalabs maybe keep this one for today’s issue?

another user also reported issues on playground and the api

edit: now o3-mini-high is stuck for over 10 minutes over at chatgpt…


still no update on the status page…

I was able to reproduce the issue in the playground using the Assistants API with o3-mini low and send a message to OpenAI.

Hope this helps and will be fixed soon.

1 Like