I am limited to four times almost every day, and after a week has passed, my GPTs (https://chat.openai.com/g/g-7cFbGmQHq-sales-lead-assistant) are finally working well. I especially hope that OpenAI can release a debugging mode with more access, for us to use for development and debugging.
2 Likes
They have limit for 40 messages for 3 hours
Yesterday I bought the subscription for chatgpt4, but after realizing how pathetic is current offer I immediately canceled subscription.
Guys you offer 40 messages for 3 hours for 20$. Good luck with this limit and price.
2 Likes
Jan 22 2024 and I’ve just joined Plus and I’ve just got the same thing. Way less than 50 messages and it still said I reached the usage cap. Beginning to feel I just got ripped off
This is nonsense! How would they feel if I change the amount I pay each month?
1 Like
I haven’t used ChatGPT for days and I’ve just received this error after creating a new chat and asking 16 questions. How this being counted?
Same with me. Got blocked after 19 questions in two hours of use. Clearly the service is not under normal behavior.
me too. just when i described what i want 
1 Like
OMG!
I paid 20$ per month!!!
Just For 20 message/hours??
2 Likes
Forget about 40 it capped at 5 messages, It this what $20 service looks like ?
I also have this problem. I paid 22€ for this???
“You have reached your current GPT-4 usage limit, please try again after 14:38” at 12:42
during the work - without warning, information about the limit. What’s going on ?!?!?! I’m paying for it. Netflix, MS Office, Spotify or Figma (or any other subcription that I use) do not turn off during or in the middle of work, movies or songs.
You are crazy. What’s wrong with your business model ? Just disable free use or restrict free users, not people who have been paying for the service since the beginning of the service.
OpenAI, “You have reached your current annoyance limit, please try and listen to paid users” - after 17:00 on 26 Jan 24, if not resolved to increase unusable limits then I am cancelling my Plus subscription and suggest others do the same.
My time is valuable - so when on a paid subscription and trying to create a GPT while repeating prompts to correct and only getting 30mins of use to build, I get blocked - I can not sit here and wait an hour or two with no pay for my time when contracting.
6 Likes
I was lucky enough to get 16 prompts in this morning before being jailed. At least, though said with some smirk, the 16 prompts produced results today vs a trend of hitting limits because of “network failures”.
I don’t want to abandon the pipeline of non-profit work I’m doing on the platform, but s private LLM with agents is REALLY starting to look tempting.
I just got same error after 40 minutes of using GPT-4.
Where I can read about limits for my subscription plan? I use web version of gpt and haven’t seen any info about limits while subscribing process.
Awful!
These limits are insultingly low for paying custumers. It’s not like we accumulate tokens during our downtime, is it now… the tokens get spent exactly when we are trying to get work done and we don’t accumulate them every 3 hours. If OpenAI wants to incentivize professional usage of their platform, this is a very bad way of going about it.
I stopped paying GPT+ a few months ago due to the weird bugs I was getting. Today, I decided to pay again to see how it improved - and such a dissapointment! I just counted my messages with GPT and I reached the limit after only 19 messages.
When I’m switching between 3.5 and 4 models, it says 40 messages cap… which is still not much per 3 hours. But being capped after 19 messages feels like a kick in the balls after paying them the outraging price for the service that seems to be useless with this limit.
For example, I was excited about custom GPT’s and so I tried the logo generator from their “recomended” section. It ate most of those 19 messages just by asking questions and the results are no good. You add a few refinements and suddenly you are out of limit “mid-work”…
If the limit acumulated by 20 messages every 3 hours, it might be fine (as my low usage hours would cover the hours when I actually need to work with it) - but paying such money for a tool that will just hit me with a limit warning after just a few messages seems dumb. That limit feels more like a free trial version, not something to be paid for.
For now, it seems like the free Bing GPT-4 is a better option for the money.
I’m seriously considering asking for refund.
1 Like
I have read that GPTs usage is set to 25 per 3 hours in the thread over here.
Maybe this is what you are hitting. Your “GPTs” limit (25) is less than GPT-4 (40) for the 3 hour rolling window. Not sure how the quota algorithm blends the two, or if you simply have to switch away from GPTs and go with GPT-4 and hit 40 there, then you are left with 3.5 (\infty) until your quota bucket(s) gets filled with additional requests.
From using the API, they fill the bucket continuously, and you don’t have to wait another 3 hours to use it. You may only have to wait 10 minutes, depending on your usage pattern.
For example, if you use it relatively evenly, there is not much of a delay. If you use it all in one big bang, all at once thing, you will have to wait the 3 hours. So usage pattern dependent.
This is done to even out the load on the AI servers. As the situation improves (more servers, more efficient models, optimized model designs–like the Turbo series) you will see the caps go up.
Or go with the API and go crazy. The quotas in the API are based on usage Tier (based on past payments made), but are generous and virtually unlimited, especially as your Tier level gets upgraded over time, from Tier 1 to Tier 5.
2 Likes
Please clarify this. Is it 40 messages sent by the user or 40 messages in total including those from GPT-4? Currently I am capped after only 20 messages sent (prompts).
I purchased the paid version of GPT 4, but I have encountered some limitations that I wasn’t expecting. Is this normal?
1 Like