Embedding Inference Server
FastAPI MiniLM embeddings with async dynamic batching. Server status lives on this page. Swagger is the interactive API explorer.
Server status
Live snapshot from GET /health.
loading
Endpoints
| Method | Path | What it does |
|---|---|---|
| GET | / | This landing page |
| GET | /health | Readiness probe. This panel; JSON at /health (200 ready, else 503). |
| GET | /metrics | Unused JSON alias of the /health snapshot (always 200). |
| POST | /embed | Embed texts. Returns vectors plus ttft_ms and ttfr_ms. |
| GET | /docs | Swagger UI generated from the FastAPI OpenAPI schema. |
| GET | /redoc | ReDoc view of the same schema. |
| GET | /openapi.json | Raw OpenAPI document. |
Try POST /embed
TTFT
-
enqueue to encode start
TTFR
-
enqueue to embeddings ready
texts
-
dim
curl
curl -X POST /embed \
-H "Content-Type: application/json" \
-d '{"texts": ["hello world"]}'