Where is this vague?
People avoid complaints like yours like the plague because the people bearing them usually just want to fight rather than face the facts.
Did anyone here on the plus plan not get a few focused coding sessions a week?
And as far as the rest… people can only explain to you that you have a weekly limit, as well as a limit of how much codex you can push every 5 hours so many times before it becomes blatantly apparent that you’re just looking for for a reason to complain and using the lack of understanding as a shield to do so.
The weekly limit has practically nothing to do with the 5-hour limit, but when the weekly limit is used up, you have spent all of your credits.
They have a pricing page for people so that they can know exactly what they’re getting
But you still feel like you need an admin to tell you all the stuff that has already been told to you.
There are 5 million people using Codex each month, and just a small few who are obstinate about understanding the limit of credits and the limit of how intensely you can spend those credits.
Should OpenAI stop everything they’re doing so you guys are out of the millions of users who clearly understand the documentation, how it works etc?
Nope.
One guy took it out of his day to help both of you understand a very simple thing but instead of learning you’re still here complaining about it.
Did you even know everyone’s codex limits were reset yesterday?
We reached 5 million users apparently, and now everyone’s got some fresh time…
I really didn’t expect to see you here complaining about this since everyone got a bunch of free codex time.
The irony here is that you keep acting like people are “too stupid to understand limits,” while completely ignoring the actual point being discssed.
We understand there are weekly limits. We understand there are 5h intensity limits. We understand compute costs money. You repeating the same thing 15 times does not suddenly address the criticism people are making about transparency and misleading presentation.
If the UI says “5h remaining,” a normal paying customer will obviously assume they still have meaningful usable access remaining. That is basic human expectation, not “refusing to understand documentation.”
You keep pointing at pricing pages and token systems as if every normal Plus subscriber is supposed to study OpenAI infrastructure economics before using a feature advertised inside ChatGPT Plus. Most people are judging the experience based on the actual product interface they are given — and right now that experience feels confusing and inconsistent for many users doing serious repository work.
And honestly, the “millions of users understand it” argument is weak. Millions of users barely touch advanced Codex workflows at all. Heavy repository users are the ones stress-testing the system hard enough to expose where the UX breaks down. That feedback matters whether you personally like hearing it or not.
Also, acting condescending toward paying users because they are frustrated with a product they actively WANT to use is such a weird attitude. People are upset precisely because the tool is genuinely useful and they want it to feel reliable instead of constantly running into invisible walls.
And no, a temporary free reset does not magically erase the criticism either. Giving users extra credits for a day does not suddenly make the communication around limits clear. It just delays the exact same frustration until the reset runs out again.
You keep framing criticism as “people wanting to fight,” but honestly the only aggressively hostile person in this thread has been you.
At the end of the day, this entire thread could’ve been avoided if the system was communicateed clearly from the start instead of users having to reverse-engineer how their own paid limits work through trial and error.
The fact that multiple paying users independently arrived at the exact same frustration should already tell you this is not some isolated “people refusing to understand” situation. When many users misunderstand the same thing, that usually points to a comunication or UX problem — not mass stupidity.
And honestly, the constnt condescending attitude toward users giving feedback is part of why these discussions become heated in the first place. People are not angry because Codex is bad. They’re angry because the product is GOOD ENOUGH to become part of their real workflow, then suddenly unpredictable restrictions slam the brakes on everything with confusing messaging on top of it.
That creates frustration. Naturally.
Nobody here demanded infnite free compute. Nobody said OpenAI should “stop everything” for them. People asked for:
clearer communication
more transparnt usage reporting
more predictable limits
fallback access after exhaustion
and UI woding that matches reality
That is completely reasonable feedback for a paid product.
And honestly, if the response to that feedback is basically:
“shut up, read documentation, millions of users don’t complain”
then you’re not defending the product anymore — you’re defending bad UX and dismissing legitimate user experience issues for no reason.
Anyway, I’ve said everything I needed to say. Hopefully OpenAI improves the transprency around this because clearly a lot of paying users are running into the same confusion and frustration.
Zamin, don’t waste your time on him. Seriously.
This isn’t a discussion — it’s someone defending a product instead of listening to real user experience. When there are no counter-arguments, “5 million users understand it fine” shows up. That’s not an argument, that’s a subject change.
I opened Codex today — it worked for about an hour, then stopped. The 5-hour window exists, but real working time inside it is one hour max. On a serious project where the model just reads files and burns the limit before doing anything useful. That’s paying for the oven to warm up, not for the food. The product is sold at full price — which means every paying customer has every right to critique it as a finished product. This isn’t a complaint thread — it’s feedback they’re obligated to hear.
Leave him to talk to himself.
You keep acting like the existence of documentation automatically means the UX cannot be criticized. That makes no sense.
A product can have documentation and STILL communicte things poorly inside the actual interface. Those two things are not mutually exclusive.
If large numbers of users are arriving at the same confusion independently, the problem is probably not “everyone is stupid except me.” Usually that points to unclear design, unclear wording, or expectations created by the UI itself.
And honestly, “you could’ve asked ChatGPT to explain the limits to you” is not the defense you think it is. A user should not need to interrogate an Ai assistant just to understand whetheer “5h remaining” actually means usable remaining time or not. The interface itself should communicate that clearly.
That’s basic UX design.
You keep framing this discussion like people are emotionally incapable of accepting limits, when in reality people are criticizing how the limits are presented and experienced during actual usage. Those are completely differrent things.
Also, constantly defaulting to condesceension (“truth trigger you?”, “just read docs”) doesn’t make your argument stronger. It mostly just makes it look like you’re more interested in talking down to frustrated users than engaging with the actual criticism being made.
Trust me, if I fully explained how it feels to watch “5h remaining” sit there while Codex locks me out after one serious repo session, we’d hit the forum character limit.
Why are you giving him a reason to have this conversation? He’ll continue to generate replies in the GPT chat. It will be an endless conversation. He doesn’t think, he just writes prompts.
People keep acting like the complaint is “why do limits exist,” when the actual complaint is:
why does the product PRESENT itself one way whlle the real usage experience feels completely differrent during serious work?
And your oven example is accurate true, also the previous bicycle example. On larger repositories, a huge chunk of the allowance can disappear before you even reach the part where you’re productively working. By the time the context loads, files get scanned, instructions processed, reasoning kicks in, and the workflow stabilizes, you’ve already burned through a massive amount of usage
That’s why the “5h remaining” wording feels disconnected from reality for many users
Nobody would be this frustrated if the product was bad. People are frustrated becuse Codex is genunely useful enough that they WANT to rely on it, but then the limit system suddenly turns the experience into “carefully ration every interection or your week dies.” That’s not a great feeling for a paid productivity tool.
At this point you’re arguing against points nobody actually made because it’s easier than addressing the real criticism directly
Nobody asked to be “coddled.” Nobody said limits shouldn’t exist. Nobody said documetation shouldn’t exist. You keep responding to exaggerated versions of people’s arguments instead of the actual point being repeated over and over:
the user experience and wording around the limits feel misleading during serious usage.
And honestly, the fact you keep demanding “examples” while sitting inside a thread full of users independently describing the exact same experience is kind of ironic.
People are literally explaining:
“5h remaining” sounds like usable remaining time
heavy repository context burns usage much faster than expected
– weekly lockouts make the displayed 5h feel meaninglass afterward
the interface does not clearly communicate the real-world behavior
That IS the criticism. Repeatedly.
You personally disagreeing with it does not magically make it invalid.
Also, the “literally the full list of users” screenshot argument is weak because forum participation is not representative of total affected users. Most people don’t make forum accounts just to write essays about UX issues. Many just get frustrated silently and move on. That’s basic internet behavior.
And finally, there’s a huge differrence between “criticism standing up to scrutiny” and someone entering a thread already assuming everyone frustrated is just irrational, emotional, or incapable of understending documentation. That attitude guarantees the discussion turns hostile before it even starts.
At the end of the day, users criticizing confusing UX in a paid product is normal. What’s weird is acting personally offended every time someone says the current experience could be communicated better.
Zamin, you fell for it.
He’s not trying to understand your problem. That’s not the goal. He’s bored, has ChatGPT open, and is generating replies to troll people on the forum. The irony is beautiful — he’s burning Codex limits to defend Codex limits.
Now the facts he never managed to answer:
— Codex burns tokens, not hours
— Reading project files, loading context, processing instructions — all of this consumes tokens BEFORE the model does anything useful
— The interface shows hours — but charges through a token counter that is never explained openly
— Result: you see 68% remaining in the 5h window, try to work — the system silently blocks you and offers to sell credits. Until next week.
That is not a comprehension problem. That is a UI honesty problem.
Leave him to talk to ChatGPT. We have a screenshot — no prompt can argue with that.
No amount of gaslighting is going to change the fact that you expect more on 5 dollars a week than you already get which was the point you raised in your thread.
You keep thinking that the other number has something to do with your overall amount of credits.
The other number only has to do with how fast you can spend your remaining credits.
When you have no remaining credits, you seem to think that other number means something.
Good luck getting the help both of you need… but it probably isn’t something this forum can provide you with.
The product is sold at full price — which means every paying customer has every right to critique it as a finished product. This isn’t a complaint thread — it’s feedback they’re obligated to hear.
