# Nuvia Free Transcriber

Free Whisper transcription service running in Hugging Face Spaces.

**Endpoint:** `POST /transcribe`  

**Form-data fields:**
- `file` (binary, required) – audio file (mp3/m4a/wav)
- `model` (string, default `base.en`) – whisper model ID

**Response JSON:**
```json
{
  "text": "full transcript",
  "segments": [{"start":0.0,"end":2.1,"text":"..."}],
  "duration_sec": 418.31,
  "words": 1120,
  "wpm": 160.43,
  "model": "base.en"
}