Hey everyone!
I’m a Plus subscriber and use GPT-4o quite a bit, especially for programming help and coding tasks. But for the past 3-4 days, I’ve noticed a pretty noticeable dip in quality. The chat doesn’t follow instructions properly, returns incomplete code, and randomly changes functionalities — even when they’re explicitly forbidden in the prompt! It feels like the model has stopped “thinking” and keeps losing track of context. 
The same thing happens with non-coding tasks too. I even tested it through the API, and the drop in quality is just as bad.Anyone else experiencing this drop in quality? 
Got any tips or workarounds you’d recommend? Curious to know if it’s just me or a broader issue!
Same here. cannot read files or pictures as well