Yes.
For further details see:
https://help.openai.com/en/articles/6950777-what-is-chatgpt-plus
Note: Currently My Account should be noted as My plan and that there currently is a bug that does not allow a user to manage their ChatGPT subscribrition by getting to the associated pages.
Also the rate caps change from time to time.
This has been a massive waste of money especially since I live in a country where 20$ is not something you can just throw around.
DALL-E won’t let me edit photos due to ‘privacy concerns’ even though there’s hundreds of other AI’s that do so, such as Adobe’s. This makes it virtually useless as a photo editor.
DALL-E keeps using this absolutely terrible "AI artistic style’’ that ruins every photo and claims it’s due to privacy and to protect creators. Even though there’s hundreds AI that are very capable of using realistic styles, for free.
Then I thought I may as well use the text GPT to write some prompts for my thesis and surprise - there is a limit. A limit that is very easy to reach since I constantly have to inquire about mismatched sources and other mistakes GPT makes. Again, a waste of money.
As with any tech ‘innovation’ I can only assume that the actual product is reserved for government agencies and high bidders and us subscription paying serfs have to make do until we get something useful in 15-20 years time
It’s now 40 msgs / 3 hrs for GPT-4 and 25 msgs / 3 hrs for GPTs.
So I was thinking once you burned through your 25 for GPTs, you could then switch over to your 40 for GPT-4. So you would have 65 “premium” calls every 3 hours.
What I don’t know is if the 25 for GPTs, also counts against your 40 for GPT-4. If so, you would only be able to burn 40 max every 3 hours, and GPTs would stop first after 25 calls. So you would then have 15 GPT-4 calls after your 25 GPTs calls.
Maybe somebody in the forum can confirm if GPTs/GPT-4 are counted separately, or is it just 40 total. If separately you get the 65 total “premium” calls, otherwise you get 40 total.
PS. The reason why it would make some sense to hit the GPTs cap quicker than GPT-4, is that GPTs are doing RAG, which takes more resources. The caps are a way to limit overwhelming the servers. And also cap the cost, since ChatGPT is a fixed cost product.
If you want virtually unlimited usage/cost, you need to switch over to the API. You don’t need to be a “programmer” to use the API either because of the Playground from the API page. You just miss out on GPTs and the more polished look of ChatGPT. So this option would be best if you use ChatGPT mainly for GPT-4 and not GPTs.
I am strictly referring to GPT-4. I do not use GPTs.
If the website says 40 messages / 3 hours I expect to be able to prompt 40 times within 3 hours.
Right now this is not the case as for the past few weeks I have been consistently capped at only 20.
Please clarify if 40 messages means 40 user prompts using GPT-4 alone.
Confirmed. That is my understanding. Are you in the USA? What country?
Does country matter? Because as others have also been saying it’s capped at 20 prompts. Perhaps you should perform the test yourself and confirm.
From my experience using ChatGPT, 25 GPTs and 40 GPT-4’s per 3 hours seems about right.
Granted 99.999% of the time I use the API.
So if another forum member that uses ChatGPT more extensively could verify this, it would be good.
I need solid data before I bring this up to OpenAI directly, that’s why I am asking. I do believe there is confusion surrounding this topic that needs clarification from OpenAI.
Simply start a new chat on the web using GPT-4 with no prior chats in the last 3 hours and count your prompts before being capped within 3 hours. Ignore GPTs, APIs, etc. Just use GPT-4 at https://chat.openai.com/?model=gpt-4
If you never use GPT4 till reaching cap, why act as GPT4 guru?
Most of us that rant about 20/3 cap are talk about GPT4 or GPT Classic, which should have a 40/3 cap. (50/3 cap before)
Why 99.99% API guy answear as if having 500 hrs experince in GPT4?
You can scroll up in this Topic, to look at how many ppl expereince this issue, and you still need a solid data?
This is a problem of GPT4/Classic user that need to fix ASAP.
I’m not alway had this issue, but twice a day is too frustrating.
Many people complain this in either Reddit or here, but API guy always defend OpenAi and act like we’re stupid.
I would like to know if anyone has tried switching devices when encountering abnormal message restrictions. You might want to try using different browsers on computers, mobile phones, and other devices, as well as the app, to see if the message restrictions are consistent. Additionally, I strongly recommend that anyone experiencing this issue immediately change their Chat GPT account password, as it’s likely that there is an issue with the account. After you change your password, try sending a message to GPT-4, and the abnormal message restrictions may be lifted immediately.
Please check the security of your various accounts as well as the security of your computer. Has anyone noticed conversations in their chat history that do not belong to them, or unfamiliar languages being used? Please check your account security. I recommend frequently changing your Chat GPT account password for protection.
This is becoming unusable and will stop my subscription and would encourage everyone to do the same to re-establish a better balance.
I didn’t use GPT4 for about a month after getting that “high cap usage” (very unusual for a paying service) but ok at that time, I was asking a lot of questions so I could understand.
10mn ago, I started reusing GPT4 and litteraly just wanted to try 2 plugins: logo creator and consistent character creations. I typed 2 short sentences for each to introduce what I wanted. I answered about 5 questions each… And I got that high cap usage message !
Now I can only imagine and hope it’s because those custom GPTs are creating a lot of messages without us knowing… but come on, 2 sentences for introduction + answering 5 questions for each and that’s it ? Well I am not paying for that.
(**its suspect by the developer that the new limit may be silently 20 prompts per 3 hours…)
I never said I was a ChaGPT guru, just trying to help. 
I will send 20 - 45 prompts to ChatGPT using GPT-4 only, periodically, and see if I get an unexpected cap. And take it from there.
But the transient outages may be related to server loads, or something like that.
Not apologizing for OpenAI at all. Understand the confusion and frustration.
Hey, have you noticed that the interface is missing the plugin option? I was wondering, maybe you can try to enable the plugin from the settings, only to find that there’s no option to select plugins on there.
Thank you for your understanding, and I apologize for any harsh words used previously.
To provide more details:
I have been experiencing this issue for over a week, and I have not used GPTs during this time.
My work involves translating large excerpts as accurately as possible. Based on my experience with GPT-4 and GPT classic, providing less text results in more accurate outputs.
Consequently, I divided the text into 35 parts (out of the 40 msg quota). That is the reason why I’m always challenging with caps.
This approach worked well until about two weeks ago, which coincided with the introduction of the “team workspace.”
About why I’m not using the API, I believe that the “conversation” format gave better results in my situation.
Now, I have to divide into 17 parts, which quite downgrade quality of the translation.
Edited:
This issue is not always happen, for me who use 4 waves a day, This problem appear 1-2 times a day.
So, last 2 day, I’m not risk anymore and playsafe with 17 chats.
OK, understood @6enjiiiz . But why 17 if you hit the 20? Is there some reason you hold back the 3?
Also just ran 40 questions, and then got to 41 and hit this message:
I understand you like your translation task better with ChatGPT, but that is something the API is primed and ready for and excels at. You should really consider the API, IMO, and ask us for help if you are lost on getting that setup.
This includes the “conversation” style history sent to the model to improve the translation performance you are seeing with ChatGPT.
(PS This wouldn’t even involve programming, you would just run a conversation thread in the Playground through through the built-in web interface, just like ChatGPT, your workflow copy/pasting to a field, and interacting, hitting “Send”, is the same as ChatGPT)
Also, it took me 45 minutes to ask 40 questions. Quite exhausting!
Please try leaving a 3-hour gap and then give it a try.
From my experience, even when using GPT with ChatGPT Plus, I was able to use about 40 messages every 3 hours.
I’m not sure if this applies to all accounts, but there’s a possibility that the message limit for GPT has been relaxed to match the regular(not GPT) message limit.
This is so ridiculous. I am using a paid version, but I have limitations? What is the rationale to suspend for app for 3 hours? So frustrating. Shame on you ChatGPT !..
Why limit is not mentioned when buying plus version? This is ridiculous and misleading.
I am so tired of the limitations, why should I pay for the extended version if I then have limited access and spend most of this limited time trying to correct the AI so that it follows my suggestions correctly? Waste of money. I will unsubscribe.
