Endpoints
Full surface; see the OpenAPI spec for shapes.
| Method | Path | Description |
|---|---|---|
| POST | /v1/chat/completions | OpenAI Chat Completions (streaming default) |
| POST | /v1/completions | Legacy alias for /v1/chat/completions |
| GET | /v1/models | List models the caller is allowed to use |
| POST | /anthropic/v1/messages | Anthropic Messages drop-in (translates to / from OpenAI) |
| GET | /qgre/v1/usage_per_key | In-memory billing-of-record table |
| GET | /qgre/v1/latency | Per-route P50/P95/P99 latency |
| GET | /qgre/v1/upstreams | Upstream pool snapshot (health, in_flight, …) |
| GET | /healthz | Load-balancer probe (no auth) |
| GET | /metrics | Prometheus exposition (no auth, scrape from inside the trust boundary) |
| GET | /openapi.yaml | OpenAPI 3.1 spec |
Full request / response schemas live in the API reference (rendered from the OpenAPI YAML).