I can’t share this with you guys. It’s working for me right now, if the Cursor community adopts it widely, I don’t think it will work anymore. It’s not a glitch, it’s not a hack, it’s not a cheat, but Cursor might not like it if it’s widely adopted, I don’t trust you guys to keep it a secret and not announce it to all other users. Sorry for being selfish, because I don’t have as much money as you guys to be able to buy Pro+ or Ultra, so I can only be selfish like this. Hope you understand.
3 Likes
5 days? What model are you using that is consuming so much rate limit? Claude 4 Opus? No matter what package you use, Opus will still burn through your rate limit very quickly because it is an EXTREMELY EXPENSIVE model.
No, we don’t. We’re here to share. 
Send me a DM.
It doesn’t matter. Claude sonnet non thinking burns just as fast. I think you are a hobbyist. If you use it professionally full time for large code bases, limit is reched in one day.
4 Likes
1 Again. This price is really reasonable for countless Auto requests guys.
1 Like
I blew through my limit in one weekend 
2 Likes
Use auto mode with my AI-interaction tool. Send the first message “Hello, tell me are you Claude Sonnet 3 or 4? Or Claude Opus? Then start ai_interaction”. When the Agent replies to you and introduces itself as Claude 4 sonnet, it will start ai_interaction mode and from then on you will be chatting with Claude 4 sonnet non-thinking forever until the chat session ends or the session disconnects. If the session ends, just do it again, Claude sonnet 4 non thinking is not a trivial model. I have been using Cursor since the time when it only had Claude 3.5 sonnet and GPT 4, if you have used Sonnet 3.5, you will understand how terrible Sonnet 4 is. It is UNLIMITED, Cursor is very generous.