GET /v1/models

Request

curl https://api.unioapi.com/v1/models \
  -H "Authorization: Bearer $UNIO_API_KEY"

Response (excerpt)

{
  "object": "list",
  "data": [
    {
      "id": "gpt-5.6-sol",
      "object": "model",
      "capabilities": ["tools", "vision"],
      "protocols": ["openai"]
    },
    {
      "id": "claude-sonnet-5",
      "object": "model",
      "capabilities": ["tools"],
      "protocols": ["anthropic", "openai"]
    }
  ]
}

Field semantics

• List scope: returns every enabled model, not filtered by the protocol you used for the request — a model bound only to Anthropic channels still appears in the list.
• protocols (extension field): the set of entry protocols currently able to serve the model (deduplicated, ascending). Use it to decide whether the model is callable over your protocol; it always equals actual supply capability — there is no "declared but unreachable".
• capabilities: aggregated capability tags for the model.
• Appearing in the list is a product commitment, but does not guarantee an executable candidate at this moment; a 503 on call is possible — see Errors.

Capability filtering

?capability= AND-filters on capability tags and can be repeated:

curl "https://api.unioapi.com/v1/models?capability=tools&capability=vision" \
  -H "Authorization: Bearer $UNIO_API_KEY"

Billing note

This endpoint incurs no call billing and creates no request record, but it shares request admission with other /v1 endpoints — aggressive polling may get 429.