Rate limits
1 limit near capacity
Requests per minute
Rolling 60 second window
412 / 500 req
Tokens per minute
Input and output combined
148,000 / 400,000 tok
Concurrent requests
In flight right now
18 / 40 req
Queued batch jobs
Waiting for a worker
3 / 25 jobs
Limits rise automatically once spend clears the next threshold.