After hitting usage limits but having a lot of success using o1 and even trying o1 pro in the api I decided to upgrade to the $200 pro plan, and it was working great a few days and now its reverted to something that cant even follow instructions at all? it feels like a completely different model and I can’t actually even use it for anything useful, I tried o3-mini-high and got the same responses, for example if I say “do this like its shown in this documentation segment” it does something completely different and generic. I will probably just downgrade to the $20 plus next month, I really did think it was going to be worth if after using o1 a bit but not if it does this
I cannot say for sure, because I don’t have the money to use it. But given that other Models also experiencing Quality degradation, it’s likely because ChatGPT is experimenting with the new 4.5 models and affecting other models. I was told this is not the first time, so it might get fixed some times down the line.
1 Like
Just an update soooo I guess with pro you only get 5 messages with o1 pro a month and mine was downgrading to o1 with a display bug that was keeping it as “o1 pro” kind of deceiving..