RS Infer
Server Endpoints

/v1/models

List loaded models. Returns one `ModelCard` per loaded model with the actual execution providers used, replicas, max_len, and whether it is the default for its kind.

GET
/v1/models

List loaded models. Returns one ModelCard per loaded model with the actual execution providers used, replicas, max_len, and whether it is the default for its kind.

Response Body

application/json

curl -X GET "https://example.com/v1/models"
{  "object": "string",  "data": [    {      "id": "string",      "object": "model",      "kind": "embedding",      "source": "string",      "execution_providers": [        "string"      ],      "replicas": 0,      "max_len": 0,      "default": true    }  ]}

/health POST

Some health-check tools only send POSTs; the response is identical.

/metrics GET

Prometheus metrics. Text exposition format. Exports `rsinfer_http_requests_total` (by route + status) and `rsinfer_http_request_duration_seconds` (histogram, by route).