How to Deploy Leon with Docker: Step-by-Step Tutorial

Deploying Leon in Minutes

In this tutorial you will deploy Leon, the open-source personal AI assistant, on your own server. Leon is self-hosted, so you keep full control over your data while getting voice and text interaction, wake word detection, and a powerful skill system.

Prerequisites

Step 1 — Clone and Configure

Clone the repository and install the dependencies. Leon uses Docker as the simplest path, so no manual Python or Node setup is needed:

git clone https://github.com/leon-ai/leon.git
cd leon
cp .env.example .env

Open .env and set your language, wake word, and port. The default web UI runs on port 8080.

Step 2 — Build and Start

docker compose up -d

This builds the Leon image and starts the server, the web app, and the voice components. Wait for the logs to show that all services are ready.

Step 3 — Create Your First Skill

Open the web app at http://your-server:8080, create your user, and test a built-in skill. To add custom abilities, use Leon's skill API:

python3 leon generate skill my-skill

Each skill lives in its own folder with a clean Python handler — write the logic, restart the container, and your assistant learns a new trick.

Step 4 — Make It Secure

That's it — you now own a private, extensible AI assistant. No subscriptions, no cloud dependency, no data leaving your server.