---
title: "Health"
method: GET
path: "/health"
---

# Health

`GET /health`

## Request

- Base URL: `https://api.voicelabs.now` (the document declares no server; this is the origin it was published from)
- URL: `https://api.voicelabs.now/health`
- Auth: none declared

Health check endpoint.

## Response `200`

Successful Response

- HealthResponse: Response model for health check.
  - `status` string, required
  - `model_loaded` boolean, required
  - `model_downloaded` boolean, nullable
  - `model_size` string, nullable
  - `gpu_available` boolean, required
  - `gpu_type` string, nullable
  - `vram_used_mb` number, nullable
  - `backend_type` string, nullable
  - `backend_variant` string, nullable
  - `gpu_compatibility_warning` string, nullable
  - `readiness` string
  - `gpu_oom` GpuOomHealth: CUDA OOM counter + last event, for a dashboard or watchdog polling ``/health``.
    - `count` integer
    - `last` GpuOomLastEvent: The last CUDA out-of-memory the engine recorded (VoiceLabs, ``voicekit_gpu_oom``).
      - `at` string, required
      - `phase` string, required
      - `engine` string, required
      - `model_size` string, required
      - `requested_mb` number, nullable
      - `free_mb` number, nullable
      - `total_mb` number, nullable
      - `reserved_mb` number, nullable
      - `allocated_mb` number, nullable
      - `resident` string[]
      - `eviction_ran` boolean
      - `evicted` string[]
      - `generation_id` string, nullable

---

[API](https://skmtc.dev/voicelabs/apis/voicebox-api.md) · [All operations](https://skmtc.dev/voicelabs/apis/voicebox-api/llms.txt) · [OpenAPI document](https://skmtc.dev/voicelabs/apis/voicebox-api/revisions/a7a37d1fc02f?raw)