Rate limits

How many requests you can make, and what happens above the limit.

Limit	Default
Speech and translation requests per key	60 a minute
Speech and translation requests per organization	300 a minute
Other API requests per key or login	600 a minute
Concurrent speech and translation requests per organization	10

Above a limit you’ll get 429 rate_limited or 429 too_many_concurrent_requests, with a Retry-After header in seconds. Authenticated responses also show where you stand:

RateLimit-Limit: 60
RateLimit-Remaining: 59
RateLimit-Reset: 42

See your organization’s limits with GET /v1/rate-limits. To lower one key’s limit, for example to contain a noisy integration, use PATCH /v1/api-keys/{key_id}/limits. Contact us to raise your organization’s limits.