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

# Rate Limiting

Layer uses a single API-wide endpoint cost rate limiting system, where every endpoint contributes a cost towards a total budget. The budget follows a bucket model, where each API call uses tokens from a current bucket which is refilled every 1 second.

## Default Rate Limits

Most endpoints use the following default rate limits:

| Environment | Limit        | Refill Period | Initial Size |
| ----------- | ------------ | ------------- | ------------ |
| Sandbox     | 100 requests | 1 second      | 100 requests |
| Production  | 500 requests | 1 second      | 100 requests |

Some endpoints have custom rate limit policies with different limits. These are documented on the individual endpoint pages.

## Rate Limit Headers

Rate limits should never be approached for normal customer read operations, and are only intended to limit heavy data ingestion processes, such as en-masse backfills of data. Nonetheless, we recommend adding retries on all Layer API calls. To assist in retries, the following headers are included in every successful API request:

* `X-RateLimit-Limit`: a specified bucket capacity.
* `X-RateLimit-Remaining`: the number of tokens remaining in a bucket.
* `X-RateLimit-Reset`: a UTC timestamp (in seconds) that specifies the time of refilling a bucket.

Any API calls that are rate limited will receive a response with:

* Error code `429: Too Many Requests`
* `Retry-After` header indicating (in seconds) how long to wait before retrying the request.

During normal operations, adding enough retries that requests will retry 1 second later should be sufficient to avoid any rate limiting.