9 Likes
I’m running into the exact same issue since the latest update. It only occurs with 3.7-sonnet(not MAX) atm, it just won’t process any requests.
Edit: I’m being billed for those requests as well.
4 Likes
I have the same issue as of this morning.
I woke up and it’s not working.
Same here
i have the same issue too, it will say it will do something and then it will not do anythign at all…and why are these counted as calls when its not working…
3 Likes
this is the part I don’t like.. I lost over 150 fast requests trying to code the same simple components over and over.. trying to fix simple console errors OVER and OVER.. each time it was charging because it defaults to fast requests in chat. Im definitely sending them another email.. I expect they will take care of everyone who’s been effected.
I wouldn’t trust it, I’ve been emailing since the start of the month and they refuse to even respond to emails anymore.
It does seem fixed to me. It was really bad earlier today, but this evening it’s been working like it’s old self.
1 Like
Claude is losing its ability to be a great helper.
2 Likes
Hey, it seems like we had some issues under the hood, unfortunately!
All appears to be resolved now, and we are looking into this internally to ensure this doesn’t happen again.
If anyone has lost a significant number of fast requests, please drop us an email at [email protected], and we’ll be happy to assist you with compensation!
I think it has not been fixed yet. Currently, Claude 3.7 series does not work. Switching to 3.5 works, but it also takes a long time to work. 3.7 waits for less than 1 minute and then stops automatically. It does nothing. I am using 0.47.8, and 0.47.7 is the same.
I would like to add that just now, 3.5 also did not work properly. It worked to one-third and then stopped automatically. I did nothing. Now almost all Claude models have problems.
I used Claude 3.7 Sonnet but it’s more or less like an idiot savant. It’s really fast in the beginning but does nothing afterwards unless you instruct it with another query. ChatGPT I think acts pretty much the same way.
I use Claude 3.5 Sonnet. This model to me is the cat’s meow, an undisputed champ. It writes really clean code, slows down when you want it to and it does have a sense of humor!
One more tip I want to give you is to write out your .cursorrule fully. You can tell it to do things in a methodical way instead of just spitting out code which always leads to a huge mess!
2 Likes
Cursor should stop relying heavily on Claude and explore alternatives. Claude keeps getting dumber due to resource limitations
At this point it’s been 3 weeks and they haven’t even responded to my emails. Did everyone quit or get fired at cursor or something? I’m having the same issue where it just responds with statements and refuses to write code. This is a series of outputs after restarting the composer and only letting it have context around the file we’re editing.
I’m not an expert here, but maybe you can try to clear cache, open a new chat, and re-index your files.
Restarted chat, re-sync index, and this was the result of the first attempt with a single relevant file in context as you can see. however I’m not upset with cursor, this has to be Anthropic, I just did a quick trial of windsurf giving it the same input and it resulted in the same outcome, both are using claude 3.5. So I think this has to do with Anthropic responses.
1 Like
Have the same problems since 2-3 days. I write a prompt and can see “generating…” for about 30 minutes without response. I already tried different models, reinstalled Cursor etc. Nothing helped. But I can see too they still charge everything.
1 Like