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

# GET /api/models

> Cached and rate-limited model catalog proxy for OpenRouter models.

# GET /api/models

Returns model catalog JSON from OpenRouter, with cache and abuse protection.

## Behavior

* Upstash Redis cache key: `openchat:models`.
* Cache TTL: 4 hours.
* IP rate limit: sliding window (30 requests / 60s).
* Requires trustworthy client IP extraction using `TRUST_PROXY` mode.

## Responses

* `200`: JSON model payload.
* `400`: Client IP unavailable when rate limiting is enabled.
* `429`: Rate limit exceeded.
* `502`: Upstream OpenRouter failure.
* `503`: Service unavailable when fail-closed infra settings trigger.