I just hit the cap limit after 20 messages over the course of about an hour this morning. I fully understand the current need to have a cap, but even 50 is pretty limiting. Why would it cut off after only 20?
Well, same here. I just hit the cap limit at 15 messages (a few were to one of my GPT).
I think, I will cancel my sub for now until it gets solved. Perplexity or the API key are also doing their job for me.
//edit: cancelled.
I did not even have used for long now i used only for 4-5 messages and then i got message that i reached a cap how i did not even send 30 message still why solve this bug or give our money back
It’s still a problem. I’ve been capped again after only 11 messages. Hello admins!
same question. got only about 20 messages 
Canceled the suscription. Pretty intransparent not to tell users that there is a limitation.
Opened Chat GPT this morning (haven’t used it for several days) and wasn’t working. Just asked me to regenerate each time. I went to my phone and tryed the app too, same thing. Non explanation, just “regenerate”. Than I noticed that Chat GPT 3.5 was actually working (not 4, though, which is disappointing). Than I tried talking to her on the phone and she finaly gave me a clue: You have reached the current message cap, try later. What the heck, I didn’t do nothing for days, and I have the paid version. Can anyone explain what’s happening?
I have an browser extension to count the number of queries made to ChatGPT and I just hit the limit at exactly 20 prompts. That was while using Dall-e, so no custom GPT. Seems like the limit went from 40 to 20.
2 Likes
I’m experiencing the same issue and also I’d like to know whats the point of creating on GPT when you can’t actually build the gpt in the first place without waiting 2 hours every time to access. In addition there is no way any user can use your GPT without being on a paid subscription. I’m not going to bother now its a total waste of time! Why put effort into training a GPT then it just goes to feed the chat gpt model which they make money from?
I’m capped now as well by configuring a GPT.
The thing is, I see the GPT builder deteriorating - and fast: I started using GPTs since its first availability, and I had a break from November until today.
It’s amazing how SLOW the GPT system is, and I couldn’t even finish a basic GPT test and then was capped.
I’m not here to complain or diminish the service, but I can’t imagine my enterprise customers trying to access their data via a custom GPT that’s slow, breaking, and capping this way.
This needs more attention and care from the OpenAI team.
It seems to be back to 40 now. Seems like they lower the limit during peak time. If they do, they should advertise as such.
So it definitely is the case that you’ll get less messages when using GPTs?
Could you learn anything about how the messages are counted? Like, do I have 40 msgs for GPT-4 AND 25 for GPTs? How does this work?
I tried a few GPTs just now, and I was capped pretty quickly. I have not counted, but I’d assume that 20 - 25 messages is accurate.
Though, I am also locked out of GPT-4. So I did not get 40 messages for GPT-4 here.
What would happen if I just do like 24 messages on GPTs, would I still have 26 messages on GPT-4 left?
This is incredibly confusing and frustrating. OpenAI should really be more transparent about this.
I mean, if it stays that way, that would suck.
But it sucks even more to just “randomly” be capped in the middle of your workflow. I’d at least like to know the specific rules behind all of this, so that I can calculate accordingly.
However, if messages sent to GPTs count as double the amount (Because I think it kinda looks like thats the case), using them is not really practical.
if we have a cap on subscription. then its not useful anymore. i cancel my subscription
I really hope the sheer number of complaints in this thread helps the developers recognise the need to priorities this issue. Just signed up with the intent on creating my own GPT to streamline activities, and got half way through training it before hitting this limitation. Absolutely nuts.
A large part of the premise of paying for the service is so that the free version can remain free for everyone else, and yet having used the free 3.5 version for half a year or so, I have not once encountered a cap on use.
I’m only glad that the ability to pay for an annual subscription doesn’t seem to be an option, as I’d be feeling even more angry than I am right now if so.
So frustrating!
1 Like
Just did the same thing. I ended up asking for a refund through Apple. Hopefully I’ll get the money back and I’ll never buy this again until they remove the limit.
This is still happening in February 2024, I barely use it, and today I just around 10 question and it throw that alert, it says try again at 11:45 AM. I am like wait so I am paying and I get limited access?
same here, february 2024 still happening
1 Like
Same here, did about ten or eleven messages, then says I can only try again after 11:30am. Which is annoying as I am paying 20 dollars a month. Now it’s only eight messages before it caps for three hours sometimes. Ridiculous.
Ive been a plus subscriber for like 6 months. First time I have seen the cap of gpt4 this morning. I have had some very heavy usage prior to today and never saw a cap. I sent maybe 10 messages this morning and I’m capped for 3 hours. I gotta think theres something wrong happening. This isn’t normal
Did Open AI ever bother responding to this or are they too busy counting their $$ and how to make more with new plans - that will probably be capped anyway.