I used to be able to get 60m tokens usage out of the $20 plan, but now I get 100% at 23m tokens?
Same for the $60 plan, where i used to get 160m tokens usage but now it gets to 100% at around 60m tokens usage..?
Why are these changes not clearly made visible/stated?
I should of also mentioned that I am using OPUS 5 and the information above was provided using OPUS 5 through Cursor.
Cursor quietly lowered the quota for other models. Originally, with Pro you could use $45 of premium model quota, and with Auto you could use $150 of quota! Now, the premium model quota is only $11–$22, with only 10–20 million tokens! Meanwhile, Auto has $450 of quota! Is this forcing us to use the Grok model? Then why wouldn’t we just buy Grok directly instead of using Cursor?
**Since being acquired by Musk, Cursor has become less and less cost-effective! The quota used to last a long time, but since the acquisition, it gets consumed extremely fast — and now the quota has been directly cut in half! Is Cursor even usable anymore? Many people on our side have started abandoning Cursor and switching to alternatives — Muse seems pretty good too! If you keep going like this, honestly, your loyal users will all leave! Just like your own company team — quite a few people have already left too!
I’ve attached screenshots below — you can look at the usage history. Basically, even a medium-level GPT can only handle about 14 messages before it’s basically done!**
Previously, regardless of task difficulty, we could still get dozens of messages. Now even medium only allows 10-something messages! A serious reduction! I’m paying $20, and I can only use $11 worth of Claude or GPT? Have you ever considered that some people don’t use Auto or Grok at all?
It looks like we’ll have no choice but to go use Claude or Codex
Right now, one Pro+ premium model quota equals one Pro — meaning you’re paying $60 to buy what used to be a single Pro
