HA TTS endpoint requires 'tts.piper' not 'piper'
Tested successfully:
- POST /v1/audio/speech → HTTP 200
- Returns 28KB MP3 (96 kbps, 22.05 kHz, mono)
- Audio: 'Alexander! I LOVE YOU!' in Tomoko's voice 💖
6.8 KiB
6.8 KiB