Rate limits
Rate limits are restrictions that our API imposes on the number of times an API key can call endpoints within a specified period of time.
We measure rate limits in the following ways:
- Requests per minute (RPM): The number of requests sent in the past minute.
We also limit concurrent generations for some endpoints. When maximum concurrent generations have been exceeded, the request will still be accepted, however it will be queued and completed sequentially.
By default, you’ll be allocated a Standard Instance, which is essentially computational resources shared among multiple API customers. If your service has higher usage volume and requires higher rate limits, concurrent generations, and faster generation speeds, you may be eligible for a Dynamic or Dedicated Instance. Please contact [email protected] to learn more.
Standard rate limits
| Endpoint | RPM |
|---|---|
| Text-to-video | 50 |
| Image-to-video | 50 |
| Text-to-audio | 50 |
| Text-to-music | 50 |
Standard concurrent generations
| Endpoint | Concurrent |
|---|---|
| Text-to-video | 10 |
| Image-to-video | 10 |
| Text-to-audio | 20 |
| Text-to-music | 20 |