GET /v1/models

Returns every model reachable on your tier, drawn from (1) every BYOK provider you have configured plus (2) the catalog entries from the internal llmMetadata table that match those providers.

How the list is built

  1. Each configured BYOK provider contributes its primary modelName.
  2. The catalog (llmMetadata) expands each configured provider with the broader inventory the SDK can call.
  3. Every entry is gated through the central authorization choke point — entries you cannot call (geopolitical block, model denylist, host policy) are silently omitted, mirroring OpenAI's "if you can't call it, you don't see it" semantics.
  4. Model ids that collide across providers are prefixed <provider>/<model> so SDK models.list stays unambiguous.

Response

{
  "object": "list",
  "data": [
    {"id": "gpt-4o-mini",            "object": "model", "created": 1714200000, "owned_by": "openai"},
    {"id": "claude-3-5-sonnet-latest","object": "model", "created": 1714200000, "owned_by": "anthropic"},
    {"id": "aws-bedrock/anthropic.claude-3-5-sonnet-20241022-v2:0",
     "object": "model", "created": 1714200000, "owned_by": "aws-bedrock"}
  ]
}

Notes

  • This endpoint never persists payload bytes — only metadata (modelCount, status, latency) is logged.
  • Returns an empty data: [] when the request is unauthenticated or no providers are configured.