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

# Rate limits

> Throttling expectations for REST and guidance for WebSocket clients.

## REST

Endpoints may enforce **per-IP** or **per-key** rate limits. When you exceed the limit, the API typically responds with **429 Too Many Requests**.

**What to do**

* Honor **`Retry-After`** when the response includes it.
* Use **exponential backoff** with jitter for retries.
* Prefer **WebSocket streams** for high-frequency market or account updates instead of polling REST.

Exact quotas are environment-specific and may change — monitor `429` rates in your metrics and reduce concurrency or batching when you approach limits.

## WebSockets

WebSockets are suited for continuous updates. Still:

* Avoid **flooding** subscribe/unsubscribe or chatty application-level traffic.
* Respond to **`Ping`** / keepalive requirements so the server does not drop you for idleness.

If you are disconnected after a burst of messages, slow down or verify you are not leaking duplicate subscriptions.
