Be careful sir. OpenAI Instead of fixing those issues or atleast learn from Mistakes by mentioning their competitor, they gonna flag your comments under excuse “Spam: Promoting irrelevant Products”. I mean i used to mentioned that google Gemini products to comparison on how ChatGPT devolved severly and give them heavy Critique based from my user expierience.. only to found I got flagged and my comments removed.
A month later and it’s gotten even worse. Go ask chatgpt what new token limits we now face without using their API, even on paid plans. Reading your uploaded docs does not truncate, but its responses get truncated now programmatically which completely breaks any kind of code rendering that breaks their super low token limit on the AI’s replies. Sometimes a continue button appears, and sometimes it doesn’t. When it doesn’t this leads to the belief that the AI has finished rendering your code block where in reality it is just sitting there incomplete.
Token limits on replies also applies to files that the AI can render for direct download, so your “full file” that you download will get truncated as well.
The new ChatGPT version 5 is a complete loser. I had absolutely no problems with version 4.
Version 4 was a completely reliable AI assistant in terms of understanding commands, remembering them, and executing them without overwriting previous instructions, learning, and acting appropriately.
However, the moment it upgraded to version 5, it feels as if it regressed from a 30-year-old adult to a 5-year-old child. It doesn’t understand instructions, gets confused by commands, and overwrites them.
Compared to the new Gemini upgrade, GPT-5 is a total failure. I hope someone from OpenAI reads this, although I’m not sure they even care.
I completely concur with your assessment. After September 15th, Codex has become an unusable tool.
Did you end up going over to Gemini? I’ve been thinking of trying it out
The way I rationalized this problem was this…
“I would use the first general release reasoning agent that came out (o1), to help save the world…”
Well that was 15 years ago…
Google had a lot of wins in that time but never released… (I have my sources)
Now I know OpenAI has some issues… That is why I come here as an API developer…
Teething Problems…
My AI can have any attitude I design it for.
Yes, I started using it more often now
That’s awesome, keep up to date I say, wouldn’t want anyone missing the point.
Yeah, there is something seriously wrong with 5.1 pro or whatever lately. it’s horrific now. forgets everything, feels like conversations stall out faster, thinking times out/takes forever compared to before, they are trolling honestly.
GPT 5.1 is a Flagship model? what a sick joke! It was worst than it’s predecessors
Not our experience. Are you sure you know how to use it?
why does openai block every post
I Don’t know, man I Think OpenAi Seems keep blocking any comment who giving negative feedbacks about Their Services. how funny instead addressing real problems they decide to Censor any critiques and ranting about their what’s wrong with them. Well I recently just got 6 Flag due my long rant comment about how poor ChatGPT Service is especially after recent update and boom! it’s Gone! with various Illogicial Reason from “Spam” to “Innapropriate”
How Ironic OpenAi Claimed to Support “Free Speech” and yet they have censorship system that worst than Authoritarian Government system
Who’s “Our”? Who do you talking to? I’m not sure if you either the one who not properly use it or just ignore every criticism that people keep talking about,Well if you’re being so obessed about it then tell me about your experience. We’re all friends around here, right? then spill it.
Well if you’re being so obessed about it then tell me about your experience.
I’m not the one who is obessed. Maybe if you provide more detail about what you are having a problem with, someone could help you. Afterall, that is the intent of this forum.
I have had failures over the last year with a few OpenAI APIs. But I didn’t give up. Failure is an important learning experience.
BTW, “Our” is the dev team I work with. Also, I just realized that this thread is about ChatGPT - I never use ChatGPT - my mistake.
But Sir.. people Are saying if this is such huge downgrade than previous model.. it’s Easily to forgets everything It doesn’t understand instructions, gets confused by commands, and overwrites them. and in Rare case. it give me double generation from Same Topic. Not to count sometimes “Project” feature is so horrible it rarely using the well established document i upload and being overstepping by creating something i don’t even need yet. beside why only point the finger at me if bunch of people in same comment are having the same thought about it.
Edit:AAAAHHHH!!! It makes sense now 
I don’t see your end of statement too!. I Thought you’re Dev of ChatGPT.. My Apology too! Have A Great Day!
So we are cool now.
Since you are a storyteller, check this out:
I told him that I’m not satisfied anymore and that GPT-4.1 was much better. I also asked him for advice since I’m going away.
His answer was: “I’m really sorry. I recommend trying Claude because it’s better than GPT-5, and I will update OpenAI about this matter”
hhhhhhhhhhhhhhhhhhh
Absolutely Sir! The Blame is on me and I’d Apologize for that
It was My Bad tho And Yeah I’ve Check your Thread sir… And Thank you for the Tips Sir.. I’ll Try it out!