Talk to AI with Voice Mode

Voice mode is one of those features that feels like magic the first time you try it. Instead of typing every command and reading every response, you can just… talk. Hermes listens, thinks, and speaks back. This guide walks you through everything you need to get voice mode running, from a simple microphone setup to a full Discord voice channel bot.

What voice mode is good for

Voice mode shines when your hands are busy or you just want a more natural way to interact with AI. Here are the most common use cases:

Hands-free CLI workflow — keep coding or researching while you talk to Hermes
Spoken responses in Telegram or Discord — get voice replies alongside normal text messages
Live Discord voice channel — have Hermes join a voice channel for real-time conversation
Quick idea capture — dictate thoughts while walking around instead of typing

Three ways to use voice mode

Hermes offers three distinct voice experiences. Pick the one that fits your needs:

Mode	Best for	Platform
Interactive microphone loop	Personal hands-free use while coding	CLI
Voice replies in chat	Spoken responses with normal messaging	Telegram, Discord
Live voice channel bot	Group or personal live conversation	Discord voice channels

A good path is to start simple: get text working first, then enable voice replies, and finally move to Discord voice channels if you want the full experience.

Step 1: Make sure text mode works first

Before touching voice mode, verify that Hermes starts and your provider is configured. Run hermes and ask something simple like:

What tools do you have available?

If that’s not solid yet, fix text mode first. Voice mode builds on top of it.

Step 2: Install the right extras

Depending on which voice experience you want, install the corresponding extras:

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 (optional):

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

Local NeuTTS (optional):

python -m pip install -U neutts[all]

Everything:

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

Step 3: Install system dependencies

Voice mode needs a few system packages to handle audio input, conversion, and playback.

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 each one matters:

portaudio → microphone input and playback for CLI voice mode
ffmpeg → audio conversion for TTS and messaging delivery
opus → Discord voice codec support
espeak-ng → phonemizer backend for NeuTTS

Step 4: Choose your speech providers

Hermes supports both local and cloud speech stacks. The easiest and cheapest setup uses local STT and free Edge TTS:

STT provider: local
TTS provider: edge

Add any cloud keys to ~/.hermes/.env:

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

# Premium TTS (optional)
ELEVENLABS_API_KEY=***

Speech-to-text recommendations:

local → best default for privacy and zero cost
groq → very fast cloud transcription
openai → good paid fallback

Text-to-speech recommendations:

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

Step 5: Recommended configuration

Here’s a solid conservative 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"

The submit_mode setting controls what happens after transcription:

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

For editable voice drafts, set submit_mode: "draft".

Using voice mode in the CLI

Start Hermes and turn on voice mode:

hermes

Then inside the CLI:

/voice on

Recording flow:

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

Useful commands:

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

Great CLI workflows

Walk-up debugging: Say “I keep getting a docker permission error. Help me debug it.” Then continue hands-free with follow-ups like “Read the last error again” or “Now give me the exact fix.”

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

Accessibility: If typing is inconvenient, voice mode is one of the fastest ways to stay in the full Hermes loop.

Tuning CLI behavior

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

voice:
  silence_threshold: 200
  silence_duration: 3.0

Lower the threshold if it stops too early, or raise it if background noise keeps it recording. The duration controls how long of a pause triggers the stop.

Voice mode is a game-changer once you get it set up. Start with the basic CLI loop, then expand to messaging platforms and Discord voice channels as you get comfortable. Happy talking!