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

# Rate limits

> Limits by tier, per key, project, and organization, and how to read the headers.

Inference is limited by your organization's **tier**. Limits apply at three scopes at once: each API key, each project (1.5× the key limit), and the organization (2×). Tokens per day are limited at the organization only.

| Tier | Reached when lifetime payments reach | Requests / min | Requests / day | Tokens / min | Tokens / day (org) | Concurrent requests (key / project / org) |
| - | - | - | - | - | - | - |
| Free | — | 60 | 10,000 | 1,000,000 | 1 billion | 4 / 6 / 8 |
| Tier 1 | \$20 | 120 | 15,000 | 4,000,000 | 2 billion | 8 / 10 / 12 |
| Tier 2 | \$50 | 300 | 25,000 | 8,000,000 | 5 billion | 12 / 14 / 16 |
| Tier 3 | \$200 | 500 | 50,000 | 12,000,000 | 10 billion | 16 / 16 / 16 |
| Tier 4 | \$1,000 | 700 | 75,000 | 16,000,000 | 20 billion | 32 / 32 / 32 |
| Tier 5 | \$5,000 | 900 | 100,000 | 20,000,000 | 50 billion | 32 / 32 / 32 |

Requests, tokens per minute, and requests per day use sliding windows. Your tier rises automatically as you pay; a plan can also grant a tier.

## How tokens are counted

A request is admitted with an estimate: its input text (about 4 bytes per token) plus the output it may generate (`max_tokens`, or the model's default). When it finishes, the counters are corrected to the actual usage. A request bigger than a whole window's budget is still admitted into an empty window.

## Headers

Every inference response carries OpenAI-style headers:

| Header | Example |
| - | - |
| `x-ratelimit-limit-requests` / `-tokens` | `60` / `1000000` |
| `x-ratelimit-remaining-requests` / `-tokens` | `59` / `998512` |
| `x-ratelimit-reset-requests` / `-tokens` | `1s`, `1m3s` (durations) |

and IETF headers, such as `RateLimit-Policy: "rpm";q=60;w=60, "tpm";q=1000000;w=60` and `RateLimit: "rpm";r=59;t=1`. `count_tokens` sends only the `rpm` entries. When several scopes apply, the headers report the one closest to its limit.

## When you hit a limit

You get `429 rate_limited` with a `Retry-After` header in seconds, and a message naming the scope and metric. The SDKs wait and retry for you. Spread bursts out, or move traffic to another key or project if one scope is the bottleneck.

## Other endpoints

| Endpoints | Limit |
| - | - |
| `POST /v1/messages/count_tokens` | Counts toward requests per minute only |
| `GET /v1/credits` | 60 / min per key, 300 / min per organization |
| `GET` on `/v1/models` and the Policy and Admin APIs | 120 / min per credential, 600 / min per organization, 10 concurrent per credential |
| Other Policy and Admin API calls | 60 / min per credential, 300 / min per organization |


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.