# termux-stt
> Android On-Device Unified STT Framework for Termux

## Core Features
- Unified create_engine() interface for whisper.cpp, vosk, and sherpa-onnx
- Built-in speaker diarization via pure-Python K-Means and Vosk 128d X-Vectors
- Zero external ML dependencies (no numpy, no sklearn, no torch)
- Subprocess crash isolation and Android mobile safeguards (WakeLock, Doze bypass)
- Live Audio Showcase: https://uno-km.github.io/termux-stt/showcase.html

## Python Quickstart (3 Lines)
```python
from termux_stt import create_engine

# STT only
engine = create_engine("whisper", model="base", lang="en")
result = engine.transcribe("audio.wav")
print(result.text)

# STT + Speaker Diarization
engine = create_engine("hybrid", lang="ko", num_speakers=2)
result = engine.diarize("meeting.wav")
for seg in result.segments:
    print(f"[{seg.speaker}] {seg.text}")
```

## Node.js Quickstart
```javascript
const { createEngine } = require("termux-stt");
const engine = createEngine("whisper", { model: "base", lang: "ko" });
const result = await engine.transcribe("audio.wav");
console.log(result.text);
```