What is the request limit of cursor pro?Unlimit?
if Unlimit, What is the different between pro plan and ultra plan?
Here’s the announcement. Seems like pro plan is now rate-limited? However, I don’t like it that they hide the count before rate limit. Who knows I’m going to get rate limited way before I fill up my 500 token quota?
What do you guys think?
Same here, I don’t like hiding the usage either. I’m not interested in unlimited usage, since 500 requests per month is enough for me. But now, with things like this, I feel a bit dissatisfied.
I just saw this too, how did it suddenly become unlimited?
same question
Hey! Under the hood we count compute used like OAI/Anthropic does for their plans.
I’m not interested in unlimited usage, since 500 requests per month is enough for me
You’re free to switch back to the old system if you’d like!
Where exactly can I change this? I couldn’t find it in the account settings or Cursor settings.
Pro Plan Changes
- Previously: 500 fast requests/month (then slow, or pay-as-you-go).
- Now (default): Unlimited requests with rate limits.
- Instead of a hard cap at 500, usage is governed by internal rate limits (e.g., requests per minute/hour/day).
- You can still opt in to the old 500-fast-requests quota model if you prefer predictable tracking.
Is it right?
I really like this change, but I hope that if possible, I can see my current usage.
such as my hourly quota, how much I have used so far, and more importantly, how much I have used, so that I can optimize my usage for different scenarios.
For simple tasks, I can use a relatively cheaper model like Gemini 2.5 Flash. In fact, this reply was translated by me using Gemini 2.5 Flash.
Where do I set it? I entered a slow request after using more than 200 requests, unable to request successfully and unable to work
I am more worried here about shady practices forcing you to make new requests.
For example, my code is written by myself. 500 requests works for me, and I am not using it every day. But when I do use it, especially after latest upgrade, Cursor borderlines with useless (for example, I spent yesterday 10 requests telling Cursor not to do something, using Claude 4 Sonnet, where it was answering with you are totally right, I apologise, and doing exactly the same. Ten times in a row. Then I just gave up. But I spent at least 10 requests this way, and probably got myself close to being rate limited without even being aware of it. And more this is happening I am less using it, and I am discovering more introductions of something that will make my user experience even worse.
Another request stealer which is becoming more and more annoying is “add domain model here, here are 29 different examples, or consistent, and add API there”, and then it finishes with adding domain model, and patiently waits for you to give it yes to continue doing what you already told it to do, but for request extra.
I am totally aware that this is not commodity, but I think everyone would appreciate transparency and not hiding counters and inventing new road blockers on the fly for people who paid year in advance. Someone really should pay some attention to UX or users will just leave, not from protests, but because it will become useless.
Any documents about rate-limited? Also, I dont see setting to change to old 500 req.
I couldn’t find a function button to switch back to the old version, and your unexpected update made me plan to use up 500 requests in a month. Just now, when dealing with a problem, I used 60 requests. If I switch back to the old version, it has already deducted 60 requests, which greatly affects my Pro experience. Do you have any solutions for this?
(post deleted by author)
says there is no limit on the number of times you can use it, but after 500 times, you will still not be able to use claude-4.
Swapping a predictable quota for a vague, “black box” rate limit doesn’t feel like an upgrade. You mentioned we could switch back to the old plan, but there’s no way to do it. How can we trust a core productivity tool when its limits are unpredictable?
Some Feedback on Ultra and The Pricing Page and Docs
I don’t know how to feel about this new plan because honestly I don’t understand it. It’s not for the want of trying. I use Cursor day to day I’ve been covering it on my youtube channel for over a year why is it so hard for me to understand the pricing? As a user this makes me feel like A. Cursor is bad at documentation or B. Cursor is deliberately hiding pricing behind complexity.
- It says unlimited everything on the 20USD Pro account. This would leave me to believe I can make as many requests as I want with as many tokens and with no rate limits across any models. There needs to be another line here explaining separate costs around max modes, rate limits
- The 20x plan highlights more usage with the models but 20 times what?
Essentially I want a clear understanding of
- What is unlimited
- How do rate limits work
- What is not unlimited and how is that priced
- A pricing case study beyond what is in the docs would be great
As a nice to have it would be great to see usage and cost appear in the chat window like Claude Code or Other Tools.
I understand there are strategic decisions being made here and Cursor is working on it’s own models and pricing is strategic. There is room for that and for clarity at the same time. Claude Code is gaining a lot of traction and the real pull is it’s 200 price plan that gets you clear Max access and rates. I’d like Cursor to be able to claim the same but right now I can’t educate myself or anyone else based on the current docs and pricing page.
Help a guy out.
Hello everyone
I’d like to say that it lacks professionalism. The site should be updated with documentation and, above all, concrete information should be provided.
How many hours are we limited to? Does slow mode still exist? Then why isn’t Claude 4 still available?
We’ve got the maximum context for the moves, so we need to update the tables and tooltips.
Anyway, it all seems to have gone out the wrong way. Going from $20 to $200 with no offer in between???
You put for the ultra mode “X20 usage” but X20 compared to what???
Now all my requests keep timing out. it doesn t work at all
Honestly, this update isn’t useful if the stability issue isn’t fixed. The app is not stabe