It was Sonnet 4, then it hits limits. Tried Gemini Pro 2.5. Still very slow.
1 Like
I did over a hundred requests yesterday without hitting limits. Are you maxing your context size with each request.
I’m a firm believer in fairness, this is a new feature. There are blind to be bugs, rather ripe in @danperks or @deanrie to see if they can investigate or log a bug.
You’ve hit the rate limit for your Pro plan. Upgrade to Ultra for near unlimited use, set a Spend Limit for overages, or switch to auto
I worked for 1 hour and then it gave me this warning. Now I can’t do anything. How is it going to happen like this? Engineers can be employed for 200 dollars in some countries.
But please, a few hours later, could you reuse it?
Come on！Baby！
Aaaaaaaaaaaa!
Is there a solution to this problem now? !
Which Contries？The Engineers are so cheap？！
Tracking for usage and plan requests is horrendous. I had, for a full month, requests coming off my balance but still being charged for usage costs.
Raised a ticket, got your AI help desk Sam, and then ghosted since…
In my opinion, I have experienced this only once; it happened on days I did not use as many requests as before, but what I noticed I had done differently was that my context window was larger, and I had used Claude 3.7 thinking… I use Claude 4 thinking sonnet more frequently.
I hit my limit around 10:40pm. I switched to other models, and I could make requests. The next day around 9am, I noticed Claude 4 Sonnet was back, and I could send requests again.
- You can try opening new windows more often
- Reduce how much context data or files you attach to your requests
- Try switching to claude 3.5 for the time being to process your requests
- Be more conscious of the prompts you send
- Make your requests in sequence; handle just a handful of task at a time
So far, so good; this may be a better option to 500 requests if rate limit is not abused. Perhaps when we learn about other people’s experiences, we will know better.
1 Like
##Update
Aparently, the cursor-4-sonnet only worked for 2 hours with 14 requests being sent with cursor-4-sonnet; at this point, it maxed out, I switched back to other gpt models
The cursor-4-sonnet model was re-enabled around 7pm, and I could send 10 requests before maxing out again.
I am currently back to using OpenAI models.
##Summary
I was able to do my work just fine; I just had to choose my models more wisely depending on what I was working on… I would say that this is fair for the time being. Let’s see what could be done better or what better options await us.
1 Like
so the only solution is $200?
Me too. The hert breaken person as a lover of claude -4
