STT

A FastAPI-based service for real-time speech-to-text using faster-whisper and WebRTC VAD.

stt_preview.mov

Installation

# Install system requirements
sudo apt install portaudio19-dev

# Install python dependencies
python3 src/setup.py
source src/stt-venv/bin/activate

Usage

Start the service:

cd src/
python app.py

Python example:

import requests

with open("audio.wav", "rb") as f:
    response = requests.post(
        "http://localhost:47102/transcribe",
        files={"file": f}
    )
    print(response.json()["text"])

Endpoints

Method	Path	Description
GET	/health	Check service health and loaded model
POST	/transcribe	Transcribe audio, with optional segments, word timestamps, or translation
POST	/vad/analyze	Analyze uploaded audio for voice activity
GET	/vad/status	Check VAD availability
WebSocket	/ws/vad	Real-time voice activity detection
WebSocket	/ws/stt	Streaming speech-to-text