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

# Usage Limits

We set no request allowance and no monthly cap. Every limit on this page is a ceiling on how much runs at once.

## On your own keys

Only one limit applies. Your organization can run a fixed number of flex races at a time. Past that number we run the request on the standard tier, so you lose the discount, not the answer.

## On Managed Keys

* **Concurrency.** Your organization can run a fixed number of managed requests at a time. Past it you get [`429 rate_limit_exceeded`](/errors#rate_limit_exceeded) with a `Retry-After`.
* **Shared pools.** Workers AI and Foundry serve everyone from one pool, so each organization gets a per-minute rate there. Past it you get the same `429`.

## On every organization

We count requests per minute three ways (per organization, per key, and per IP). A fourth counter tracks failed auth attempts per IP. A body over 50 MB returns [`413 request_too_large`](/errors#request_too_large).

Back off for the seconds in `Retry-After`.

`GET /v1/limits` reports every ceiling on this page next to the spending ones. See [Spending Limits](/spending-limits#the-limits-endpoint).