I’ve been using cursor for a while now, 500 requests per month is enough to speak of.
However, I found out that in the 0.47 update, the number of requests for the thinking model will be doubled, which may cost me double or be slower if I exceed the number. Is there a version 0.46 that I can return to avoid consuming double tokens?
Or are there any official plans to increase the amount of high speed requests per month? After all, cursor is no longer considered in the cheap echelon of similar editors
I’m facing the same issue. Even for someone like me who works part-time, a mere 500 requests barely meet the monthly quota. Considering the Cursor team’s policy of “IT’S JUST AS IT IS,” I suppose these are the company’s choices to maximize profits. Perhaps it would be better to reconsider using Cursor or even start writing my own Cursor!
Regarding your question, I believe the version of the client doesn’t impact the usage calculations on the server-side.