You've reached the current usage cap for GPT-4, please try again after 2:04 PM

I have started today with only 10 messages and generating two pictures. then i got a message that said I’d reached the cap usage. after just 10 prompts?!! so disappointing that OpenAI sells the Plus version with these limitations. 10 messages every 3 hours? is it different for every user?

2 Likes

I suspect that internal chats that the model has with itself to complete the request eat up the final number. For example, when you ask for an image, gpt rewrites your prompt to make it more “complete” and that probably counts as at least 2 messages with this scheme. I believe this is the case because I notice my usage is consumed faster when I ask for things like that instead of stuff that gets answered with a single response.

Of course, it is BS. It should only count the user’s original messages!

1 Like

I hope something changes soon. $20 is a little more than fair and I now find out there is a limit and that limit seemed to be very short.

I’ve just gotten my subscription to GPT Pro, and I am appalled by the cap. I did not know that this subscription would be limited in such a way and for $20/month that is very disappointing. I am hoping that in the future there will not be a limit to this service.

Same here, I subscribed at 11:00 AM, and at 12:47 AM I received the same message :frowning:

Bought the GPT-4 Subscription at 17:53 and got the message “You’ve reached the current usage cap for GPT-4, please try again after 8:50 PM.” after like 10 requests in Data Analysis with providing PDF’s.

Same, I want to request a refund but there’s no way to contact anyone. Paying 20 bucks to have something work for a few minutes is a rip off.

1 Like
  1. How to cancel ChatGPT Plus subscription:

https://help.openai.com/en/articles/7232927-how-do-i-cancel-my-chatgpt-plus-subscription

  1. How to request ChatGPT Plus refund:

https://help.openai.com/en/articles/7232895-how-do-i-request-a-refund-for-chatgpt-plus

1 Like

Ran into this cap after uploading about 5 photos and asking chatgpt to describe certain things in the photo. This is ridiculous.

1 Like

I was under the impression GPT4 was unlimited.
I am a newly subscriber and really quickly I was informed I reached the msg cap.

I checked online and there is something about gpt cap being 100 msgs.
That is so weird! Why would you give unlimited service to people using your free version and cap the usage of those paying for it?

2 Likes

I’m not sure OpenAI will suffer from our pitty $20 per month as they get investments from microsoft, elon musk, and so on.

It’s just sad that I paying for this service every month and should suffer from limits.

“You’ve reached the current usage cap for GPT-4”

Wtf is that ? We are paying for a service and have limits with ChatGPT ?

1 Like

Perhaps we can make a petition for the limit to be removed?

2 Likes

I’ve read over a good amount of the entries here and I couldn’t agree more, not only with the frustration of undisclosed and constantly erratic caps, but some of the specific issues which seem unbelievable that the system is constrained in its ability, developed by people that are claiming to lead the world in developing artificial intelligence that can’t even change the color of the coat in a picture as one person mentioned without changing the whole picture despite the users explicit instructions. I could go on forever with the problems I’ve experienced here and in other systems based on Chat GPT… Here’s a summary I sent to a friend of mine who asked about my experience at first regarding artwork but I expanded beyond that in my response. By the way this was condensed from my original information by about 2/3 buy another AI chat system. A lot of the details and texture, the nitty gritty was left out. But it did capture the essence quite nicely while the full impact is lost by the details which I kept for reference. I think this is really what’s behind the problems- greed and the public’s gullibility which the developers rely on to make their money.
"I’m not interested in the artistic ability of the AI, but I’m frustrated by its inconsistency and limitations. For example, it can’t make the man’s hair gray or the man petting the reindeer as I asked. It also ignores or misunderstands some of my instructions when I try to correct it. It seems to have a fixed set of components that don’t work well together.

However, it can also be very accurate sometimes, like the picture I sent you. But the artwork is not my focus. I’m more into factual inquiries, and I’m disappointed by most of the AI chats, such as ChatGPT 3.5, Bing Chat with GPT 4.0, and Bard from Google. They have significant problems and make bizarro mistakes. The only one that is pretty good at answering my questions is ChatGPT 4.0, but it also has its limitations. I paid $20/month for it, but I feel like I was baited and switched.

These AI systems are not ready for primetime. They are still in mid-development and they make too many errors. The developers rushed them out to gain market lead, but they are not reliable or trustworthy.
The AI can make huge mistakes in their answers (not always obvious). They can contradict themselves, refuse to correct themselves, or repeat the same mistake after apologizing. They can also copy what I tell them as the correct answer and pretend they found it themselves. For example, I told one of the AI that its population figure for Los Angeles was wrong and ranked it 4th in the state of CA. It apologized, gave me another wrong number, and repeated my phrase about LA’s ranking word for word. It’s incomprehensible, because it’s a machine learning program with databases that were far from ready for prime time, but greed and disrespect for the customer prevailed.

I think the high tech world has been unethical and craven, especially with the coming of AI. Bill Gates started this with his use of words like “bug” and “glitch” to make the public think these things just happen when developing “leading edge” technology. He did this to avoid proper beta testing and roll out software long before it was ready, making the customers suffer and pay for it. He made an obscene fortune this way while we still suffer. He then became the richest person in the world.

The current “AI Guys” took it to a new level and rolled out these products even earlier in the development process, just to be the leader. These products were far from ready and still in development stages.

I can tell you more about how I really feel if you like.

Despite all of that, I have found some of my queries useful and time-saving, if I understand the limitations and possibilities. Other queries have led me to wrong answers.
Some of the AI’s answers were wrong or partial and I wasted hours trying to figure them out. It can’t correct itself well. Chat GPT4.0 seems better, but I need more time with it.

That’s a brief summary of my reply.

Thank you and I’d like to hear your experience too."

1 Like

Canceled my subscription too 20 min reached cap for my game with customgpt

Was looking forward to setting up my own GPT to work with, and was well on my way to setting it up when I got the message.

For me the biggest issue is that it’s a total stop: no downgrade to 3.5 option; no option to turn off image generation or data analysis, no Super Plus subscription where you pay extra for no limit - nothing.

Given this I’ll probably drop my Plus subscription. Custom GPT is practically useless for anything beyond the most shallow use cases, and it was the only feature that seemed worth paying for.

1 Like

I just want to say it here for anyone who feels tricked by OpenAI.
I contacted Apple requesting them to refund me the money I paid for ChatGPT Plus right after realising the company was capping the amount of msgs and I am glad to say they approved my refund.
As a customer I feel like there was lack of clarity and I do not understand why they’d give unlimited msgs for non-paying users and cap the amounts of msgs for those who pay.

To say I’m frustrated is an understatement; I had been practicing my prompt before I decided to pay for Plus, and most of my very clear instructions were not considered. For example, portrait (vertical) vs landscape (horizontal). Don’t get me started on full body image, even with the shoes being described in detail.

I did NOT know there would be a cap when I considered GPT-4 for Dalle 3 purposes. I decided to go with ChatGPT vs the other options because I also wanted the ability for the richer content prompts as well, but in all honestly I was fine with version 3, which doesn’t have a cap.

I may use this for the end of the subscription and go to a different avenue.

Does anyone know if the usage cap serves a functional purpose? At this point, I’m just curious as to why OpenAI would implement such a thing.