> ## Documentation Index
> Fetch the complete documentation index at: https://docs.ekly.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Rate limits

> What the limits are, how to read the headers, and how to behave on a 429.

| Limit | Scope |
| - | - |
| 60 requests per minute | per API key |
| 10 generation submissions per minute | per organization, across all keys |

Reads (models, estimates, polling) count toward the per-key limit only.

## Headers

Every `/v1` response carries:

```
X-RateLimit-Limit: 60
X-RateLimit-Remaining: 57
X-RateLimit-Reset: 1759485660
```

`X-RateLimit-Reset` is a Unix timestamp for when the window clears.

## On 429

The body has `code: rate_limited` and the response carries `Retry-After` in seconds. Sleep that
long, then retry. Polling every 2 seconds stays well inside the limit; if you run many generations
at once, poll them in one loop rather than one loop each.

Need more? Write to [hello@ekly.ai](mailto:hello@ekly.ai) with what you are building.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.