Has the usage cap for ChatGPT-4 been reduced or do custom GPT generations count as extra uses?

I finally was able to sit down and use my newly created custom GPT after all the network issues seemed to have been resolved, and I thought I’d be able to generate around 30 outputs before taking a break, but I was only able to generate 23 outputs before I hit the usage cap. So I’m wondering if the cap has been reduced or if using my custom GPT to generate outputs costs more. I haven’t seen anyone else complain about it so far, so it could also be a bug (not the first time that’s happened to me if that’s the case.)


Happened again. Looks like its reduced to 25 uses every 3 hours again for me. Before I send a bug report, is anyone else experiencing it?


Could somebody please just tell me if they are able to generate more than 25 ChatGPT-4 outputs every three hours? I’m not asking for much here.

1 Like

Same problem here. Working on a GPT and was able to generate 22 outputs with it before reaching the usage cap, now have to wait another 2,5 hours before I have access again. Seems like the cap is reduced or GPTs cost more in some way…


At least I know it’s not just me then. I’ll wait for more replies before I contact support (which from my experience doesn’t help much anyway).

Will try generating with default ChatGPT-4 later to test it.

1 Like

OK, I’ve been using the regular ChatGPT-4 now, and managed to get 36 generations so far, and I think that’s enough to know it’s functioning as normal. That means the issue is only with custom GPTs.

I guess I can live with that as I rarely need more than 25 generations in a 3 hour period, but I wonder why it’s happening. My main issue is that if each custom GPT generation really counts as 2 (or more) uses, it will look like I’m using GPT-4 more than I actually am, and I’ll unfairly be seen as a very heavy user.

Edit: Looks like that ChatGPT-4 and the custom GPTs have separate usage caps, so you can actually get 73-75 generations every three hours. Probably shouldn’t get used to it though, as it seems like an oversight.

1 Like

This is also happening to me, I’m getting 24 or 25 generations with my custom GPT. I use a plugin for GPT to keep track of messages, and it says I still have 16 messages left, but the chat says I’ve run out of messages. This didn’t happen when I used standard GPT 4 the other day.

Yeah, it seems to be 23-24 for custom GPTs, and you can use the remainder of the (now) 40 generations per 3 hour period on regular ChatGPT-4.

usage is temporary reduced and i hope the raise the usage cap to a usable level.

Have made 10 GPTs using and try and error process and the result is i wast rapidly all my usage.

Adding to this to also say I’m experiencing the same. I hope this is only temporary because at this rate, it’s difficult to actually put the GPTs to proper use.

1 Like