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

# Rate limits

> How Zenamu throttles API requests.

The API allows **100 requests per minute**, per endpoint.

Because the counter is per endpoint, calls to `/v1/classes` do not consume the
budget of `/v1/workshops`. A page that loads classes, workshops, lecturers, and
places gets 100 calls per minute on each.

Treat 100/min per endpoint as the ceiling for your whole integration, and design
to stay well under it.

Exceeding the limit returns `429`:

```json theme={null}
{
  "statusCode": 429,
  "error": "Too Many Requests",
  "message": "Rate limit exceeded: up to 100 requests per minute per endpoint."
}
```

Every response includes these headers:

| Header                  | Meaning                                            |
| ----------------------- | -------------------------------------------------- |
| `X-RateLimit-Limit`     | Maximum requests in the current one-minute window. |
| `X-RateLimit-Remaining` | Requests remaining in the current window.          |
| `X-RateLimit-Reset`     | Unix timestamp in seconds when the window resets.  |

Use `X-RateLimit-Reset` when scheduling a retry after `429`.

## Staying under the limit

**Cache reference data.** Lecturers and places change rarely. Fetch them once
per page load — or once per hour — instead of once per event.

**Request wider ranges.** One call for a whole month costs the same budget as
one call for a single day. The range is capped at 366 days, so a long backfill
is a handful of calls, not hundreds.

**Skip capacity when you can.** `includeCapacity` is more expensive. Ask for it
only where you actually show live availability.

**Use bulk grants.** [Bulk credits and passes](/guides/granting-credits) handle
100 clients in one request instead of 100.

## Handling 429

Back off and retry. The window is one minute, so a short exponential backoff
starting around a second recovers quickly.

Read endpoints are safe to retry. For the credit and pass endpoints, a `429`
means the request never reached the grant logic, so retrying is safe there too —
but never retry blindly on a `500`, which may have applied.