Replicate · Rate Limits

Replicate Rate Limits

Replicate API rate limits per account.

4 Limits Throttle: 429
Rate LimitingML Inference

Limits

Predictions create (default) account
predictions_per_second · second
10
Predictions create (paid raised) account
predictions_per_second · second
100
Other endpoints account
requests_per_second · second
60
Concurrent predictions account
concurrent
varies

Sources