3 Likes
Same here
Just receive this when using Sonnet 4
I’m seeing the same issue with Vertex—requests are crawling along. Even the 3.7 Sonnet model feels painfully slow. It looks like Cursor is intentionally throttling non-fast requests (see this Reddit thread for details).
At this point, I’m switching over to Cline or Windsurf.
Error: “We’ve hit a rate limit with Vertex”
1 Like
same, sonnet 4 thinking
1 Like
I got this too and still got charged fast request even though it said Errored, not charged?
Same here, with different models… Very frustrating!
yes also happening to me and it’s been very annoying
I’m just seeing this now.
We’re experiencing high demand for Claude 4 Sonnet right now. Please switch to the ‘auto-select’ model, another model, or try again in a few moments.
Seeing the same, can this be consider an outtage? Please help us out Cursor team.
It’s happening also with Gemini models (Flash). Not only Sonnet.
Hey, we’ve been seeing some weird behaviour from Vertex which is causing this error, and are now managing it the best we can while we work with Google on this!
2 Likes
I faced with the same error, but when I clicked the ‘Resume’ button, I was be able to continue without problem using Claude 4.
2 Likes
Hitting this more and more frequently now 
Same error here, using Claude 4
It’s my first query in the last 5 days
1 Like
Seems to be Cursor org hitting a limit. My workaround has been to subscribe to Claude Max which gives priority during peak hours. However, I’m not sure if you can benefit from this when using the Anthropic API Key in Cursor (as I’ve been using Claude Code today.)
same here
1 Like
still happening
1 Like
Still happening
Same here.
We’ve hit a rate limit with vertex. Please switch to the ‘auto-select’ model, another model, or try again in a few moments.
I’m getting the same message quite frequently today:
We’ve hit a rate limit with vertex. Please switch to the ‘auto-select’ model, another model, or try again in a few moments.