Akool · Rate Limits

Akool Rate Limits

Akool does not publish fixed numeric per-minute request-rate limits for the OpenAPI. In practice, throughput is governed by two things: your plan's monthly credit allowance (every generation deducts credits) and your plan's concurrent-generation limit (Starter allows roughly 2-4 concurrent jobs, with higher and custom concurrency on Pro Max, Business, and Enterprise). Because generation is asynchronous - create a task, then poll by id or await a webhook - clients naturally pace themselves against the concurrency ceiling rather than a per-second cap. Bearer tokens expire and must be refreshed (error code 1101 signals an expired or invalid token).

Akool Rate Limits is the machine-readable rate-limit profile for Akool on the APIs.io network, conforming to the API Commons Rate Limits specification.

It captures 4 rate-limit definitions, measuring requests, concurrent-jobs, credits, and token.

The profile also includes 3 backoff/retry policies defined and response codes documented for success, authInvalidOrExpired, authEmpty, and accountBanned.

Tagged areas include AI Avatars, Video Generation, Generative AI, Rate Limiting, and Quotas.

4 Limits
AI AvatarsVideo GenerationGenerative AIRate LimitingQuotas

Limits

OpenAPI Request Rate account
requests
not published
No fixed numeric per-minute request-rate limit is documented for the OpenAPI.
Concurrent Generations account
concurrent-jobs
per plan (~2-4 on Starter, higher / custom on upper tiers)
The primary throttle. Asynchronous jobs run against a per-plan concurrency ceiling.
Monthly Credit Allowance account
credits
per plan
Every generation deducts credits; exhausting the allowance blocks further generation until topped up.
Token Lifetime account
token
expires
Bearer tokens from /getToken expire; refresh on error code 1101. The x-api-key method avoids token refresh.

Policies

Asynchronous Create-then-Poll
Generation endpoints return a task id immediately; clients poll the info endpoint by id or receive an encrypted webhook on completion rather than holding a long request open.
Credit Metering
Requests are metered by credits consumed per operation and per output setting (resolution, length), not by a request-count quota.
Backoff Strategy
Clients should poll with a reasonable interval and exponential backoff, and prefer webhooks over tight polling loops to reduce load.

Sources