Miscalculations in Claude 4 sonnet compute usage, or request mismanagement on behalf of cursor. Dashboard says 2000000 for sonnet, with less than 200000 for other Anthropic models and other models in general. Id like to see if anyone feels the same way. I also think that we all have a particular day where our “usage” in the dashboard seems very abnormal based on what we actually did. Ive seen a couple other people also encountering 1 single day of extreme abnormal usage incurred (on the same day as me) and then facing issues with cursor ever since. Mind you I use Cursor consistently every day, I do not stress test or abuse models or my plan for fun. Let me know if this is just me tripping too.
Today, while using the Claude 4 Sonnet model, Cursor told me that I still had 9 days left before my credits would run out — it was supposedly just an estimate.
Then, after using the Claude 4 Sonnet model just once, my access was immediately revoked and I was told my credits had run out.
Something is clearly wrong here.
ive gotten 2 Claude 4 requests every night at 9pm for the past 6 days. Today… Nothing… Also told me Id have to wait until the next month before I can use Cursor again
I’m starting to believe that Cursor is deliberately pushing us to try Auto Mode just so we can have a full-blown mental breakdown, end up hospitalized, and then, in a desperate attempt to avoid losing our sanity again, feel forced to upgrade to the Ultra plan.
If this isn’t psychological warfare disguised as a product feature, I don’t know what is.
Yeah, I wont be upgrading to pro+ or ultra anytime soon as they’ve been trying to undermine and manipulate consumer value behind their opacity to maximize shareholder value
If I keep thinking about this any longer, I might actually start believing the free version is better than Pro — and that’s saying a lot. At least you are getting service interruption by not paying.
I mean it does feel like I got the old Free versions limits on the Pro version lol.