You've hit the rate limit on this model

Hello! It’s a pleasure to greet you.

Sorry to bother you, but while using Cursor with the Claude 4 model, I received the following alert message:

“You’ve hit the rate limit on this model.
Switch to a different model, upgrade to Ultra plan for 20x higher limits on Claude / Gemini / OpenAI models, or set a Spend Limit for requests over your rate limit.”

Although I was able to continue working with other models, I became curious about how this works and whether others have experienced something similar.

Here are my questions:

  • Does this notice go away and allow me to use the model again when my Pro plan renews?
  • Is it just a matter of waiting a few hours or days for the model to become available again?
  • Does Cursor provide any official information about these limits and their reset times? I couldn’t find anything on their official page.

I’d really appreciate it if someone could share any information or experience on this topic. Thank you very much!

same questions, the limit refresh after 5 hours then after 3 prompts I got limit again and wait for another 5 hours.

Your answer worries me, hehe I use a lot of claude 4 to make 3 pront and wait another 5 hrs, today I changed model to gemini 2.5 pro and it was terrible for me hehe

2 Likes

make sure you’re not sharing account
in my end I can use claude 4 sonnet thinking max mode for many hours without limit

Hey folks. There’s burst and local rate limits, and it differs for each model. So if you hit a rate limit for one model, you can switch to another and continue using premium models.
Local rate limits refill every few hours
Burst rate limits refill more slowly
Rate limit docs here:

2 Likes

to be honest your doc is ambiguous, no clear information, how long consumed context, how long waiting, etc

1 Like

In terms of consumption, I should also reiterate that rate-limiting is tied to your compute usage. So, smaller requests against cheaper models will allow you to use Cursor for much longer periods of time without being rate-limited.

We have no clear information / visibility about those ratelimits apart from vague docs.

They are very “abstract”.

Half of the topics on the forum are now users complaining about the lack of transparency regarding this matter.

Local rate limits refill every few hours
Burst rate limits refill more slowly

what does “more slowly” mean? what does “few hours” mean.

What’s the issue about having a clear cooldown UI / timer shown to the user?

It’s not the best look as it’d be easy to think there is something else going on.

4 Likes

Thank you very much. In my case, it’s my own account, which I use on two devices: my work account during work hours, and my personal PC when I make time for my projects.

Never both at the same time.

1 Like

Thank you very much for the documentation, I will read it to understand more. Thanks for your response. The truth is, I was worried that I would no longer be able to use that model until the end of the subscription. In my case, I do notice that one model is very attached in terms of response quality for work.

But if I have to wait a bit while I use it, I have no problem. At least I can’t use it daily by killing 20 or 25 prompts. That was my rule when I had 500 credits. In my case, I didn’t even reach that in a day and I reached the limit, which is why I was surprised by the message.

Yes, this situation has been a bit frustrating. After 3 requests, I also received a warning to switch to the ultra package. It seemed better when Cursor Pro had a limit of 500.

Hi, could you report a bit more detailed what you did and how you used Cursor?

Here is my personal experience and i havent reached the rate limit in 3.5h intensive work on personal account.

with stats

Show how much compute each request is costing us and how much compute is left and when cool down timer will reset it. Its illegal to charge users on something that vague and abstract

I waited about 24hrs then got 2 requests and then rate limited again :frowning:

2 Likes

It’s the inconsistency, some get amazing usage, others get practically nothing.

3 Likes

I am also restricted, this feeling is too bad.
I have been using claude 4 thinking for development, but now there is a restriction prompt. Prompt me to switch models.
After switching to claude3 and claude3 thinking, there is still a restriction prompt. After 30 minutes, a claude3 thinking request was sent and restricted again.

i bought the Pro plan to use claude 4 sonnet, but for the past two days, i havent been able to use it properly after just two prompts i hit the rate limit—it’s basically unusable for me now!!!

In my case, I use my own account, never share, use only on one project for work. I also do not constantly using chat, but occasionally type manually, in which I would say quite “healthy”. It works well in the last 2-3 days, today I hit the limit.

Using Gemini 2.5 makes my anger uncontrollable

Gemini is USELESS