Tutorial 27: Voice Mode

Welcome back, friends! Today we’re going to talk about one of the coolest ways to use Hermes: Voice Mode. If you’ve ever wanted to chat with your AI assistant while walking around, cooking dinner, or just avoiding the keyboard, this tutorial is for you.

What is Voice Mode?

Voice Mode lets you talk to Hermes instead of typing. It’s perfect for:

Hands-free CLI workflows – code and research without touching the keyboard
Spoken responses in Telegram or Discord – get voice replies alongside normal messages
Live Discord voice channels – have a real conversation with Hermes in a voice chat
Quick idea capture – dictate thoughts while walking or brainstorming

The best part? If you use Nous Portal, you get both the LLM and text-to-speech through one OAuth login. No extra credentials needed!

Three Ways to Use Voice

Mode	Best For	Platform
Interactive microphone loop	Personal hands-free use	CLI
Voice replies in chat	Spoken responses with messaging	Telegram, Discord
Live voice channel bot	Group conversations	Discord voice channels

Pro tip: Start with text mode working first, then add voice replies, and finally try Discord voice channels if you want the full experience.

Step 1: Get Text Working First

Before diving into voice, make sure the basics work:

hermes

Then ask something simple:

What tools do you have available?

If that’s solid, you’re ready for voice!

Step 2: Install the Extras

Depending on what you need, install the right packages:

# CLI microphone + playback
cd ~/.hermes/hermes-agent && uv pip install -e ".[voice]"

# Messaging platforms (Telegram/Discord)
cd ~/.hermes/hermes-agent && uv pip install -e ".[messaging]"

# Premium ElevenLabs TTS
cd ~/.hermes/hermes-agent && uv pip install -e ".[tts-premium]"

# Everything
cd ~/.hermes/hermes-agent && uv pip install -e ".[all]"

Step 3: System Dependencies

macOS:

brew install portaudio ffmpeg opus
brew install espeak-ng

Ubuntu/Debian:

sudo apt install portaudio19-dev ffmpeg libopus0
sudo apt install espeak-ng

Here’s why these matter:

portaudio → microphone input for CLI voice
ffmpeg → audio conversion
opus → Discord voice codec
espeak-ng → phonemizer for NeuTTS

Step 4: Choose Your Providers

Easiest setup: Use local STT and free Edge TTS. Add this to ~/.hermes/.env:

# Cloud STT options (local needs no key)
GROQ_API_KEY=***
VOICE_TOOLS_OPENAI_KEY=***

# Premium TTS (optional)
ELEVENLABS_API_KEY=***

STT options:

local – best for privacy, zero cost
groq – super fast cloud transcription
openai – good paid fallback

TTS options:

edge – free and good enough for most
neutts – free local/on-device
elevenlabs – best quality
openai – good middle ground
mistral – multilingual, native Opus

Step 5: Recommended Config

Here’s a solid default for most people:

voice:
  record_key: "ctrl+b"
  submit_mode: "direct"  # TUI: direct | draft
  max_recording_seconds: 120
  auto_tts: false
  beep_enabled: true
  silence_threshold: 200
  silence_duration: 3.0

stt:
  provider: "local"
  local:
    model: "base"

tts:
  provider: "edge"
  edge:
    voice: "en-US-AriaNeural"

Note on submit_mode:

direct (default) – submits transcript immediately
draft – puts transcript in composer so you can edit before sending

For editable drafts, set submit_mode: "draft".

Using Voice in the CLI

Start Hermes and turn on voice:

hermes

Then type:

/voice on

Recording flow:

Press Ctrl+B
Speak
Wait for silence detection to stop automatically
Hermes transcribes and responds
If TTS is on, it speaks the answer
The loop restarts for continuous use

Useful commands:

/voice
/voice on
/voice off
/voice tts
/voice status

Great Voice Workflows

Walk-up debugging: Say “I keep getting a docker permission error. Help me debug it.” Then continue hands-free: “Read the last error again”, “Explain the root cause in simpler terms”, “Now give me the exact fix”.

Research/brainstorming: Perfect for walking around, dictating half-formed ideas, and asking Hermes to structure your thoughts in real time.

Accessibility: If typing is inconvenient, voice mode keeps you in the full Hermes loop.

Tuning Behavior

If Hermes starts/stops too aggressively, adjust the silence settings in your config:

voice:
  silence_threshold: 200
  silence_duration: 3.0

Lower the threshold for more sensitive detection, or increase it if it’s cutting you off.

That’s it! Voice Mode is one of those features that feels like magic once you get it running. Start simple, get comfortable, and soon you’ll be having full conversations with Hermes while doing the dishes. Happy talking! 🎤