Get Single Model Info

Returns information for a single model available to the current token, looked up by model name (OpenAI /v1/models/{model} format). When both the x-api-key and anthropic-version request headers are sent, the Anthropic format is returned (id, created_at, display_name, type).

Note: when the model does not exist, the HTTP status code is still 200 and the body is {"error": {"code": "model_not_found", ...}}. If the token is bound to specific channels, only models from those channels can be looked up. Queries are free.

Source of truth in code: router/relay-router.go (modelsRouter.GET("/:model")), controller/model.go (RetrieveModel), dto/pricing.go (OpenAIModels).

Known issue (tested 2026-09-28): models whose owned_by is custom in the GET /v1/models list (about 119/164, e.g. gpt-5.5, claude-sonnet-5, aigc-video-*, qwen3.x, grok-4.x) return model_not_found from this endpoint, while models with owned_by openai / vertex-ai / ali / deepseek etc. work normally. To determine whether a model is available, rely on the GET /v1/models list rather than this endpoint.

Last verified/modified: 2026-09-28 (API Base URL unified to https://api.crazyrouter.com; example model names follow what GET /v1/models returns); tested claude-sonnet-4-5 / gemini-2.5-pro / gpt-4o work normally; custom models such as gpt-5.5 incorrectly report model_not_found (logged, fix pending)

Provide your bearer token in the

Authorization

header when making requests to protected resources.

Example:

Authorization: Bearer ********************

{ "id": "claude-sonnet-4-5", "object": "model", "created": 1626777600, "owned_by": "vertex-ai", "supported_endpoint_types": [ "anthropic", "openai" ], "public_endpoint_types": [ "anthropic", "openai" ] }

Request

Responses