Getting Started

Jarvis is a microservice-based voice assistant you can self-host. This section walks you through installation, sending your first command, and understanding the configuration system.

Prerequisites

• Docker and Docker Compose (v2+)

For GPU-accelerated LLM inference:

• macOS: Apple Silicon (M1+) for Metal/MLX
• Linux: NVIDIA GPU with CUDA drivers and nvidia-docker

Note

Python and Git are only needed if you're installing from source.

What You'll Learn

1. Installation — Multiple install options: one-liner, npx, Docker, or from source
2. Your First Command — Register a node and send a test voice command
3. Configuration — Network modes, environment variables, and service discovery