Hermes Agent voice mode

10. July 2026 · 3 min read

Hermes Agent voice mode

Hermes supports two-way voice communication — speech-to-text and response synthesis.

CLI Voice (Ctrl+B)

# Press Ctrl+B→ Speak → Hermes transcribes and responds

Telegram Voice a Discordu

Send a voice message — Hermes transcribes it and replies automatically.

STT Providers

  • faster-whisper — local, free
  • Groq Whisper — fast, free tier
  • OpenAI Whisper — most accurate

TTS (text to speech)

# config.yamltts: provider: openai # or edge (zdarma), elevenlabs

← Back to Blog