Monthly VS Weekly Usage Restrictions

Ideas Feature Requests
1

Hello everyone,

I’m not sure if this is something that’s been brought up yet but I would say that the biggest limiter that i’ve experienced so far is the usage. I am using a couple of different platforms and some of them have weekly resets alongside 5hr limits. As a student with limited funds, I would love to use Cursor more but I fear pushing that limit closer and closer to 100% knowing i have a whole month to wait.

Is there plans to move to weekly limits? Is it all the same and it’s just a mental trick other platforms are playing.

3

Hey, thanks for the feedback. I’ll reply to your two questions.

About the “mental trick”, no, the difference is real. With a monthly window, your accumulated usage sticks around longer, and it feels more stressful than weekly cycles that reset more often. So your feeling is totally valid.

About switching to weekly limits, I don’t have any specific announcements or plans to share right now, so I can’t promise anything. But it’s a valid request, so I’ll move this thread to Feature Requests so other users can vote. That makes it easier for the team to gauge demand.

4

Hey, I feel like you’re so on-top of everything haha, thank you so much for attentiveness and quick responses. Looking forward to seeing the threat!

7

If the total monthly quota remains unchanged, I don’t want any hourly or weekly limits, as that would disrupt the development schedule too much.

8

Hey, thanks for the feedback. I get your point. For some users, more frequent hourly or weekly resets are more convenient. For others, they can break your workflow if the total monthly amount doesn’t increase.

I don’t have any specific announcements or plans to change how the limits work right now, so I won’t promise anything. But this is a helpful signal. Different reset models work better for different users, and both viewpoints matter.

9

Hello!
Like OpenAI’s Codex, I’m against weekly and monthly token usage limits.
When using it for work, there are times when I have to use a lot in a single week, and other times when I hardly use it at all.
I switched to Cursor precisely because I didn’t want OpenAI’s Codex “weekly and monthly token usage limits” to interfere with my work—but if Cursor also manages usage on a weekly basis, it will be a huge problem.

Cursor is my partner.
I’d like you to continue using the current monthly usage counter.
I’m against weekly and monthly token usage limits.

P.S.
Since this is a translation from Japanese, the phrasing might sound a bit off.

10

Hey, thanks for the feedback, and especially for laying out your workflow. That’s exactly the kind of detail that helps me see the full picture.

You and @liang_ding are basically saying the same thing: if the total monthly usage isn’t going up, more frequent resets just get in the way because usage isn’t evenly distributed week to week. That’s a totally reasonable take, and it’s the opposite of what the thread author was asking for, which is exactly why there isn’t one “right” answer here.

I don’t have any announcements or plans right now to change the current monthly model, so I won’t promise anything. The monthly counter is staying as is.

If you’re interested in an idea that could work for both sides, optional daily or weekly guardrails on top of the monthly limit, meaning the monthly total stays the same, but you can set your own warnings or caps so you don’t burn through everything in a couple of days, it’s being discussed here. Feel free to take a look and vote:

P.S. The translation looks good, your point comes through clearly.

11

Hi @deanrie

I checked the topic you linked in your previous comment. However, the pay-as-you-go model could become quite expensive for us.

Would it be possible to submit a feature request to make the current Monthly Guardrail more flexible by introducing hourly and weekly reset limits?

For example, Claude currently provides a usage limit that resets every 5 hours, along with a separate weekly limit. A similar approach would work really well here and provide much more predictable usage and cost management.

It would also be great if these limits could apply to third-party model pools, not just models such as Composer and Grok. For example, when we explicitly select a third-party model for coding instead of using Auto, it could have its own hourly/weekly allowance or shared third-party model pool.

I believe this would make the offering much more attractive for developers who frequently use third-party models while still allowing you to maintain reasonable usage controls. It could also help strengthen the product’s competitiveness and reach in the developer market.

12

Hey, thanks for writing out the scenario. That’s helpful.

You don’t need to file a separate feature request. The idea of more flexible guardrails, like hourly or weekly thresholds on top of the monthly limit, is already being discussed here, and that’s the best place to add your voice and details:

Your point about having those thresholds apply to third-party model pools too, not just included models, is a great addition to that discussion. Please add it there as a comment so everything stays in one place and it’s easier for the team to gauge demand.

On monthly vs weekly resets, the position is unchanged for now. I don’t have any announcements or plans to change the current monthly model, so I won’t promise anything. But the feedback is clear and it’s useful.

A couple quick clarifications on third-party models so we don’t mislead anyone. Pool split and limits depend on your plan and current configuration and they can change over time, so the most accurate view is in Dashboard > Usage. Pricing is here:

I won’t quote exact included numbers since they aren’t static.

13

Dean Rie, thank you for your reply.

When I listen to feedback from Cursor users, I realize that opinions are indeed divided into several groups, and I don’t think there’s a single optimal solution.
If we were to listen to one side and implement their suggestions, users would likely be split between those who are satisfied and those whose work efficiency decreases.

This might be a bit of an unusual idea, but…
When signing up for or renewing a Cursor Pro, Pro+, or Ultra subscription,
how about letting users choose between using Cursor Model and Other Models under a monthly limit, or one of two other patterns: hourly, weekly, or monthly limits?

That way, users could choose between a monthly limit or either a time-based or weekly limit, allowing the service to be tailored to their specific usage patterns, don’t you think?

In my case, I have a Pro+ subscription and use Agent Mode for C# Windows Forms, WPF, and PHP, but I only use it for 9 hours a day, Monday through Friday. Since I concentrate my usage into those 9 hours every day from Monday through Friday, the “Hourly or Weekly Limit” option would be a deal-breaker for me.
This is precisely why I stopped using OpenAI and switched to Crusor.

Crusor is my partner.
If possible—and I apologize for expressing this as my personal opinion—I would like to request that the current system remain as is.

According to information from September 30, 2026, there seems to be a discussion on X about changes to Crusor’s pricing structure and other aspects.
I can’t help but be very concerned about this :frowning:

P.S.:
This has been automatically translated from Japanese to English.
Some of the phrasing may sound a bit unnatural.
Also, there may be cases where what I’m trying to convey hasn’t been accurately translated into English.

14

Thanks for laying out your setup in such detail. 9 hours a day Mon to Fri is a lot of concentrated usage. It also shows why there isn’t one single correct answer here. What works well for one user can break another user’s workflow.

Your idea to let users choose the limit model monthly vs time-based when they sign up or renew is a helpful addition. Conceptually it’s exactly what’s being discussed in the thread about configurable guardrails on top of the monthly limit. It’d be great if you could add it there as a comment so everything stays in one place and the team can more easily gauge demand.

On the monthly model itself, my position is unchanged. I don’t have any announcements or plans to change the current monthly counter, so I can’t promise anything. That said, the feedback is clear and useful.

On the discussion on X, I don’t comment on third-party posts. We haven’t announced any product changes. If anything changes, it’ll show up in official channels like the changelog and pricing docs, not in someone else’s post. Current pricing is always here:

15

Hi team, I’d also like to see this feature, plus a small amount of extra tokens added to the Pro+ plan.

When I’m using Cursor, I always have to worry about whether I’ll have enough left for next week, so it makes me very anxious. And if I run out, I have to buy Claude to keep working. If I find Claude works well for me, then not many people will come back to Cursor. This comes down to user behavior, and 15 to 20 days is long enough to change a habit.