Introduction

The sovereign operating system for synthetic intelligence.

Luca is not a chatbot. It is a full-stack operating system designed to manage the lifecycle of synthetic intelligence on personal hardware. Unlike SaaS models where your data lives in the cloud, Luca runs local-first, ensuring absolute data sovereignty.

Local Execution

Runs Llama 3, Mistral, and other SOTA models directly on your GPU/NPU.

Memory Graph

A persistent, encrypted vector database that grows with you over time.

Installation

Luca installs as a system daemon, managing the underlying inference capabilities while exposing a standard port for client interactions.

Prerequisites

  • macOS 14.0+ (Apple Silicon) or Windows 11 (WSL2)
  • Minimum 16GB Unified Memory (RAM)

Quick Start

Get up and running with the Luca Kernel in seconds using our installer script:

BASH
curl -sL https://get.lucaos.space | bash

System Architecture

The Luca OS is composed of three primary layers designed to separate concerns between inference, memory, and interaction.

1. The Kernel

The orchestrator. Manages process allocation, model loading/unloading, and hardware resource limits.

2. The Cortex

The inference engine wrapper around `llama.cpp` and `MLX`.

Skill Building

Skills are sandboxed functions that Luca can "call" when needed.

TYPESCRIPT
export const getWeather = { name: "get_weather", description: "Get the current weather", parameters: { type: "object", properties: { city: { type: "string" } } }, handler: async ({ city }) => { // Implementation } };