An Official Way to See Usage and Rate Limits in Cursor

Ideas Feature Requests
1

Hi Cursor Team and community!

I wanted to share a thought I’ve been having as a happy customer and daily user of Cursor IDE.

While browsing the forum, I came across a discussion about a third-party extension that supposedly lets users see their usage and rate limits. However, @condor mentioned that “The 3rd party extension is not official and does not show full or correct information.” And honestly, I believe that’s true.

So, here’s my idea:

Since many of us—including me—would love to know things like our exact rate limits, usage stats, and when limits reset, wouldn’t it make sense for Cursor to offer an official feature or plugin for this?

If the third-party extension might show incorrect data, I think it would be far better (and safer) for both users and the Cursor team if there was a built-in way to access these details. It would improve transparency, trust, and overall satisfaction for paying customers.

I want to be clear that I’m not criticizing here—I genuinely love Cursor IDE. It’s an amazing tool that helps me tremendously, and I don’t want to switch to Copilot or any other alternative. I just hope we could get a bit more transparency or insight into these usage conversations.

For example, I’ve noticed that even though the Cursor Team introduced unlimited requests (which we really appreciate!), there are still quite a few people saying things like “I only sent one request and already hit a rate limit.”

Also, one thing that confuses me a bit: if there’s an Ultra plan, why does the Pro plan still exist? Other tools might offer bigger token limits (like Claude, offering 10x or 20x), but we choose Cursor because it’s simply better in many ways.

Thanks so much for all your great work and for building such an incredible IDE. I just wanted to share my thoughts in case it helps improve Cursor even more for all of us.

Example of extension for Cursor Team:

image
image387×600 27.6 KB

------------------------ Check my Other Posts ------------------------

2

Guy, everyone has been asking for a while now. Cursor, aren’t going to show because it will reveal their gamification algorithm..

3

People are all worked up about the lack of clarity around the rate limits, but it’s all very simple.

Imagine your requests as a sack of magic beans, and the AI agent as a collective of wizards. When you want to ask the wizards a question, you need to pay them with beans.

You’re not allowed to know how many beans are in the sack, and the wizard you will take an unknown number of beans based on which wizard you ask and the content of your question, and maybe how busy they are. When you run out of beans, or maybe if too many other people are asking questions and the wizards get sleepy, the smarter wizards will refuse to talk to you (though you won’t know that unless you specifically ask for them by name and they tell you no) but less-competent wizards will be happy to assist you sooner or later. Don’t worry though, because an elf will be by at some unknown point to put an undisclosed number of beans back into your sack! Also maybe there are two flavors of beans that get refilled at different intervals?

I’m confident that this should clear up all questions regarding the basic use case of this, the flagship product of a $9+ billion dollar company.

7

I’m for it :slight_smile: upvoted

8

Claude 4 Sonnet Thinking. Okay. How many request you used because it will reset on 6pm. And you’re just waiting to refill up? With rate limit transparency, People can decide they should buy Pro plan or not. Because some people can’t buy 200$ Plan monthly. So this means that if there is a Pro plan, it should also be useful. Besides, If you tryna to use Opus modal with Pro Plan and MAX Mode of course(because there is only with max mode), You will see you can’t send request over 10 or 15. I don’t want to wait like hours in middle of the work sir. That’s why people are trying to tell the Cursor Team their confusions.

With thinking and max mode rate limit filling up so much faster. If I have to choose one of the two I think everyone wanna use “Thinking” modals. Because it’s better than non-thinking about Large refactors, Complex Works/Problems etc etc. If I’m gonna use claude 4 sonnet non-thinking without boring limits Why would people give $20 a month? Cursor Must be different and good from other Agent tools.

Nowadays people making Extensions, MCP tools for use more request and they want to see exactly usage. The reason for all this is because there is no transparency here like customers want… :hugs:

9

agreed, more transparency is needed

11

Hey, actually, we’ve already received a feature request for this earlier, and I’ve
logged it in our internal tracker so the team can take a look at it.

12

Question begs to be asked - over those 11 days, did the team take a look at it?

13

I’m really disappointed that Cursor has shut down the API, so third-party extensions can no longer query usage data. Hopefully, Cursor will provide official support for this feature soon.

15

The API has been restored.

17

You guys really need to stop trying to get them to make everything visible and ‘transparent’; because what your actually going to do is remind them(and by them, I mean their investors) that this year has been the best deal of all our lives, and you know it lmao. How many of you are looking at this screenshot like “only 144? what a filthy casual” :rofl: They are trying sooo hard to be cool about it for us, but like shhhhh lol
image

18

Everyone should just use auto and shut up. They can’t tell you the limits because they change based on demand.

19

At first I thought it was a great improvement, but actually the limits are quite low. I haven’t coded with Sonnet for 48+ hours and after 5 requests (small amount of context) I got this

image
image255×141 8.46 KB

21

I would love to see how many premium requests i have made too. I had no idea it was limited to 500 until ChatGPT pointed that out. However, PLEASE do not advocate for ending Pro pricing! Those of us who do this as a hobby or solo dev cannot afford $200/month! I am lucky to even afford $20.

22

Wait, with the new pricing model, I turned off ‘Usage-Based Pricing’ and used Claude 4 Sonnet MAX at no cost. In 7 days, I made over 400 requests, and I love the new pricing model

image
image1330×673 72.3 KB

23

I experienced the same thing - it was amazing! However, suddenly my MAX requests were exhausted, and now I’m confused about how it works. I don’t understand when or how I can continue using MAX, since Cursor shows a rate limit message.

My main concern is that I don’t understand how much I’ve spent and when the limit will reset.

I eventually switched back to the old plan, but Cursor removed the request cost display and counts failed requests as billable requests. So I don’t have clarity on the old plan either.

This is very frustrating. I’m considering switching to another IDE.

24

We definitely need more transparency with this. We should be able to see if we’re about to hit any limits and not just get notified when it already is too late.

25

This is a glitch, I had it working for about 30 mins haha

26

A great solution would be for the Cursor team to introduce an official in-app dashboard that clearly displays real-time usage metrics, rate limits, reset timers, and plan-specific quotas—similar to how API platforms like OpenAI or Claude show usage data. This would eliminate confusion caused by third-party extensions and enhance trust, transparency, and user experience for both Pro and Ultra users.

27

I selected Opt Out in the dashboard/settings/account advanced settings to reverse back to the previous pricing model with 500 request limits. Comparing the previous mode, the new pricing model already costing me extra 20 dollars so far because of sticking with Claude-4-sonnet-thinking model. seems cursor sets request limitation for the single model.