Limits and Credits

Credits

Every generation consumes credits based on the selected model and parameters. Image models commonly use fixed credit pricing. Video models commonly use per-second pricing with parameter multipliers such as resolution or audio.

Credit usage is calculated server-side when a generation is created. Use GET /api/v1/models for callable models, required inputs, and accepted parameters rather than internal pricing configuration.

Concurrency

API generations share the same user-level concurrency pool as web generations. Your maximum parallel tasks depend on your subscription plan. If the account already has active pending or processing tasks, new API requests may be limited by the same plan concurrency.

Upload limits

Default upload limits are:

Model modes may define stricter limits, formats, max file counts, or max durations.

Production notes

• Generation is asynchronous; design your integration around polling.
• Pass mode explicitly in production requests.
• Store returned file URLs if your workflow needs to reuse source assets.
• Webhooks, SDKs, and idempotency keys are not part of the current public API surface.