List Models

Returns the currently callable models in OpenAI's response shape. The OpenAI SDK's model-listing call can target this endpoint directly.

Authorization

header

AuthorizationstringRequired

Bearer <your RouterOne key> A RouterOne key created in the console. If you lose it, you can view it again on the keys page.

Response

200response

The currently callable models — narrowed to what the key can call when one is sent. Alongside OpenAI's standard fields, every entry carries RouterOne extensions such as display_name, aliases, pricing, and providers.

401response

The key is missing or invalid.

What a key changes#

This endpoint does not require authentication — hitting it right after you change base_url is the first step of any integration.

When you send a valid API key, you get back the models that key can actually call: its model allowlist, source scope and pinned providers all apply here, and each model's providers[] lists only the ones it can reach, with pricing recomputed over what is left. Prices also reflect your account's discounts.

An invalid, expired or disabled key is never an error — it is treated as unauthenticated and you get the full list. So this endpoint is not an access check; the real decision happens when you make the call.

Model names#

id is the bare name, that is, the model name without the vendor prefix, such as claude-opus-5; for models without a bare name, it is the canonical ID. Relay services that sync this list can merge it with same-named models from other upstreams without model mapping.

The canonical ID, such as anthropic/claude-opus-5, is also accepted and is provided in canonical_id and aliases. The model field in responses always returns the canonical ID.

Extra fields#

Beyond the standard fields, each entry also includes the following extensions. They are not part of the official OpenAI protocol; SDKs usually ignore them, but they are available over direct HTTP calls.

pricing is the discounted starting price, so during a promotion it is the promotional price. The actual charge depends on the provider selected for the request, so check that provider's quote in providers[]. Each entry carries its own unit: per_request is priced per call and input_text per token — six orders of magnitude apart, so check the unit before converting.

Some components include a freeQuantity field: the quantity that is not charged on each request. For example, when the reference-image component has freeQuantity set to 5, the first 5 reference images in each request are free and any beyond that are charged at the component's unit price. Components without a free allowance omit the field.

Three response shapes on one path#

/v1/models is shared by all three compatible protocols. The response shape depends on the auth header used for the request:

Gemini also has its own URL, /v1beta/models — see List models (Gemini).

The official SDKs already send the appropriate auth headers, so most clients require no extra handling here.

curl https://gateway.routerone.ai/v1/models \
  -H "Authorization: Bearer $ROUTERONE_API_KEY"

{
  "object": "list",
  "data": [
    {
      "id": "claude-opus-5",
      "canonical_id": "anthropic/claude-opus-5",
      "object": "model",
      "created": 1786000000,
      "owned_by": "anthropic",
      "display_name": "Claude Opus 5",
      "context_length": 200000,
      "max_output": 64000,
      "input_modalities": [
        "text"
      ],
      "output_modalities": [
        "text"
      ],
      "capabilities": {},
      "aliases": [
        "<string>"
      ],
      "pricing": {},
      "providers": [
        {}
      ]
    }
  ]
}