Hermes Live Voice

What Hermes Live Voice does

Setup prompts for your API key, installs the Live Voice plugin and starts the gateway. In the Hermes Dashboard you open Chat, select a saved conversation and press Start voice. Automatic turn detection is on by default, and you can pause the microphone or interrupt a reply without stopping background work. A Tasks panel shows progress and retained results, Stop task ends a task, and disconnecting voice lets it keep running.

The realtime provider handles speech and turn-taking, while Hermes owns conversation history, memory, tools and task execution. The gateway persists task receipts, progress, notifications and results, so task state survives voice disconnects and gateway restarts. Running Hermes work cannot survive a Hermes Agent restart, and ambiguous outcomes stay marked unknown. Work runs one task at a time by default, and approval-required work is denied and stopped. A browser SDK and a hermes-live terminal command offer other ways to connect.

Key features

• Realtime voice chat inside the Hermes Dashboard Chat view
• OpenAI Realtime as the default provider, with Gemini Live and local Hugging Face voice as options
• Background tasks with progress, retained results and a Stop task control
• Task receipts that survive voice disconnects and gateway restarts
• Doctor and launch-check commands that print concrete fixes
• A browser SDK and a text terminal client for advanced use

When to use it

• Starting a repository inspection and test run in the background while planning a release by voice
• Reconnecting later to find the results of a task you started by voice
• Embedding voice control in a host app through the browser SDK

Who it is for: Hermes Agent users who want to talk to their agent hands-free without blocking on long-running work.

How it fits with Hermes Agent

It is built for Hermes Agent as a Hermes Dashboard plugin and gateway, and the README calls it a community integration, not an official NousResearch distribution.

How to install Hermes Live Voice

These commands are copied from the project's README. Check the repository for the latest steps before you run them.

npm install --global hermes-live-voice@next
hermes-live setup
hermes dashboard

Requirements: Hermes Agent 0.20.0 or newer for the Dashboard UI, Node.js 20+ (22 or 24 recommended) and an OpenAI API key with Realtime access; managed local voice needs Apple Silicon, uv and at least 12 GB of memory.

Note: The README describes this as the 1.3 release candidate, installed from the next tag, and the default setup needs an OpenAI API key with Realtime access.

FAQ