typervox

CLI that captures audio from the ALSA default device and streams Whisper transcription to stdout. On first run it auto-downloads a faster-whisper model via ct2rs Hub (default Systran/faster-whisper-base.en, English-only, INT8-friendly for low latency).

• Set TVX_MODEL_ID to choose a model ID for ct2rs Hub.
• Set TVX_LANG to fix the language (default en).
• Run typervox and watch recognized speech print incrementally.