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

# Rate limits

Indices uses rate limiting to maximise API and infrastructure stability for all our users. These limits cap the number of requests per second we will accept per user. Additionally, we use concurrency limits on certain endpoints to limit the number of simultaneously active requests on the given resource.

If you exceed the limits, you get 429 Too Many Requests HTTP status responses. The Indices SDKs are designed to automatically retry a request. For non-SDK users, you should implement a retry mechanism.

**Rate limit:** The global rate limit is 25 requests per second.

**Concurrency limit:** `POST /runs` is limited to 25 concurrent requests. Unlike rate limits, which reset each second, the concurrency limit counts how many requests are in progress at any given moment.