API JSON response

Shape returned by the authenticated providers API and the public live dump. Credentials are never included.

Method	Path	Auth	Body
GET	/providers	Bearer `PROVIDER_SERVICE_TOKEN`	Same as below (`items` only)
GET	/catalog.json	None (public)	`items` + `syncedAt`
GET	/events	None (SSE)	Live `catalog` events with the same JSON body

Last refresh: 2026-11-15T04:00:15.204Z

Response fields

Field	Type	Notes
items	CatalogProvider[]	Hardcoded providers, refreshed on an interval
items[].key	string	Stable id Lintas AI stores as `catalogKey`
items[].name	string	Display name
items[].baseUrl	string	OpenAI-compatible base URL for completions
items[].models	CatalogModel[]	Ids, windows, capabilities, prices
models[].id	string	Vendor model id used for chat
models[].contextWindow	number \| null	Tokens; null when unknown
models[].maxOutput	number \| null	Tokens; null when unknown
models[].capabilities	string[]	e.g. `chat`, `vision`, `reasoning`
models[].prices	object \| null	USD per 1M tokens; null when unpriced
prices.inputPerMillion	number	Live input rate
prices.outputPerMillion	number	Live output rate
prices.cacheReadPerMillion	number	Cached input read
prices.cacheWritePerMillion	number	Cache write
syncedAt	string \| null	ISO time of last successful refresh (public JSON only)

Example

{
  "items": [
    {
      "key": "tokenizer",
      "name": "Tokenizer",
      "baseUrl": "https://tokenizer.id/v1",
      "models": [
        {
          "id": "deepseek-v4.1-flash",
          "contextWindow": 1048576,
          "maxOutput": 393216,
          "capabilities": ["chat", "vision", "reasoning", "tools"],
          "prices": {
            "inputPerMillion": 0.0128248,
            "outputPerMillion": 0.0513009,
            "cacheReadPerMillion": 0.00128248,
            "cacheWritePerMillion": 0.016031
          }
        }
      ]
    }
  ],
  "syncedAt": "2026-11-14T14:00:00.000Z"
}

Open the live payload: /catalog.json