Beta docs: the shipped API is POST /api/v1/generate for gemini and sarvam. Some pages below describe planned features. See the shipped quick start.

API Reference

Models API

Every model your key can call, with supported capabilities.

Endpoint

terminal
GET https://openlink-1.onrender.com/v1/models

Lists models available across your enabled providers.

Response body

JSON
{
  "object": "list",
  "data": [
    {
      "id": "claude",
      "provider": "anthropic",
      "description": "Careful general reasoning",
      "context_window": 200000,
      "can_stream": true
    },
    {
      "id": "sarvam",
      "provider": "sarvam",
      "description": "Hindi and 22 more Indian languages, speech",
      "context_window": 8000,
      "can_stream": true
    }
  ]
}

Requesting a model

Set "model": "claude" in a chat request to name a model explicitly. With "model": "auto", the gateway picks a default per provider — the id that appears in the response tells you what actually served the request.