Has anyone experienced unusual downgrades in performance lately with the O1 and O1 Pro models? Both models are making simple errors, and when I ask which model they are, they respond with ‘4O.’ The Pro model no longer takes more than a second to think before starting to write. When I report an error, it consistently repeats the same mistake. It feels like an older version of the 4O.
4 Likes
I have the same problem, this is disgusting. o1 is working like 4o or 4o-mini, not reasoning at all…
1 Like
Same here. Did someone try to contact openai about this? its disgusting to pay 200$ and literally have the o1 models bugged
I’ve tried, but their solutions never work. Incognito mode, clear cache, change browsers. I think it has something to do with the servers.