Hermes Voice

What Hermes Voice does

Hermes Voice is a thin layer. A realtime speech provider handles the conversation, and email, calendar, contacts and other tool work goes to Hermes through an ask_hermes bridge that calls your Hermes API. An async task queue (start_task) lets the model delegate longer jobs without blocking the conversation, and results return on a natural pause or through a results-ready chip. The Lounge screen has an orb that you tap to talk, hold for push-to-talk, or leave in hands-free mode with server VAD, and you can also type.

A conversation timeline is searchable and exportable, and a task orbit shows background tasks with progress and cancel. Result cards for events, emails, links and contacts need an on-screen or spoken go-ahead before real-world actions. Access uses a URL key, with an optional multi-user mode that maps one Hermes profile to each Voice user. Per-user personas, voice choice, memory review, UI locales in English, French and Spanish, and a setup wizard are included.

Key features

• Orb-based voice control with tap, push-to-talk, hands-free and typed input
• Async task queue that delegates work to Hermes without blocking the call
• Searchable, exportable conversation timeline with tool steps and approvals
• Result cards that require a go-ahead before real-world actions
• Optional multi-user mode with one Hermes profile per user
• xAI realtime by default, OpenAI Realtime through VOICE_PROVIDER

When to use it

• Talking to your Hermes agent hands-free to check email or calendar
• Delegating a longer research task by voice and hearing the result later
• Giving each household or team member a voice persona backed by their own Hermes profile

Who it is for: Hermes Agent users who want a voice interface and have an xAI or OpenAI key with realtime access.

How it fits with Hermes Agent

Built for Hermes Agent: it needs a running Hermes with its OpenAI-compatible API enabled, typically at http://127.0.0.1:8642.

How to install Hermes Voice

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

git clone https://github.com/Cosmekaili-creator/Hermes-Voice.git
cd Hermes-Voice
npm ci
cp .env.example .env

Requirements: Node.js 20+ (24 recommended), an xAI API key with realtime access or an OpenAI key with VOICE_PROVIDER=openai, a running Hermes Agent with its API enabled, and HTTPS in production

Note: It is licensed under PolyForm Noncommercial 1.0.0, so commercial use needs a separate license from the author.

FAQ