Possible bug. GPT-4o is using Reasoning or o4 mini by default

ChatGPT Bugs
1

Not sure if anyone else has noticed this, but this morning, I just received the GPT-4.1-Mini update. However, I noticed as a result, GPT-4o now uses reasoning, even though I don’t have it enabled. Because after I hit enter when I submit a prompt. I tried everything to possibly fix it, such as explicitly saying in my custom instructions not to use reasoning and even logging out of my account and logging back but nothing has worked.

Now I don’t have reasoning enabled and 4o is set to the default model in the model options (i’m a Free user). So with that being said, is it possible that I encountered some bug that either uses reasoning by hard default or is unintentionally behaving like o4-mini?

I’m using ChatGPT in a browser by the way.

3

I have the same problem. Finally found someone with the same. Yesterday morning I started to notice new dialogues with 4o have loss in quality, the message limit has decreased, the inscription “reasoning” has appeared. To all questions the model answers that it is “o3”. When I try to regenerate the answer in 1 out of 5 cases the answer came from 4o, in the rest of the cases it is o3. It’s terrible and so frustrating.