Monthly usage quota

In the past, when we used the program, there was a limit of 500 and when that limit was exhausted, we could continue to use those models, provided that it was slow, but now, because of this monthly quota, our work is left unfinished. You need to fix this problem.

Our project is left unfinished and we have to wait for the monthly quota to expire or agree to pay a higher fee. This is impossible. If it continues like this, I won’t consider renewing my membership because the old system was much better, and when the quota was reached, we could still complete the project with slow requests instead of fast requests.

Hi @skymirco and thank you for your feedback.

Please have a look at your consumption in Dashboard > Usage. It lists the models with token amount used. This depends a lot on how much context you use in chat (attached files, rules, MCPs,…).

Therefore less context used gives you more AI usage per month. Heavier context usage consumes more resources and may result in faster consumption of the plan.

While for most users the Pro plan is enough, some users with heavier coding usage requested larger plans like Pro+ and Ultra.

Overall it is possible to use Cursor for a lot of coding with good management of context.

1
I understand you very well, my friend, but the old system wasn’t like this. I’m a veteran user. There wasn’t much need for calculations back then because after fast request was over, we could still use models with slow request. That’s what I’m talking about. The old system was better for this reason. If this isn’t fixed, I think there will be people like me who think a lot about it and will cancel their subscriptions for this reason. Admins need to consider this.

2 Likes

Yes, now most AI tools are switching from requests to token based billing as a request could be few tokens or a million. While average user may not use up that many tokens, heavy coding does need good context management as the tokens add up.

Cursor Team does review feedback and considers it.

Happy to share ideas to reduce context and token usage. If you can tell me a bit how you use Cursor and how you code with AI I may be able to suggest improvements. Let me know.