# Rate limits

Pace requests and handle throttling using the response headers.

Product: Account API
Guides follow API reference 0.3.0 and event reference 0.2.0.
Canonical page: https://docs.axiym.io/account-api/integration-guide/fundamentals/rate-limits

The Axiym API platform limits request volume to protect service stability. Requests
that exceed available capacity receive `429 Too Many Requests`.

Limits are applied per source IP address as seen by Axiym.

| Limit                  | Current value            |
| ---------------------- | ------------------------ |
| Sustained request rate | `20` requests per second |
| Short burst capacity   | `100` requests           |

If several services share one NAT gateway or proxy, they share the same
capacity. Do not hard-code these values.

## Response headers

| Header                  | Meaning                                                    |
| ----------------------- | ---------------------------------------------------------- |
| `x-ratelimit-limit`     | Capacity for the current source IP.                        |
| `x-ratelimit-remaining` | Capacity remaining when the response was generated.        |
| `x-ratelimit-after`     | Seconds until capacity becomes available after throttling. |
| `Retry-After`           | Standard retry delay returned with `429`.                  |

When throttled, wait for `Retry-After`, reduce concurrency, and add exponential
backoff with jitter if throttling continues. Reuse the original
`Idempotency-Key` when retrying the same state-changing operation.