What is Mai

MAI is a local LLM client that provides AI assistance through a desktop chat application, ensuring user privacy by processing all data locally without external server interaction.

Use cases

Use cases for MAI include managing personal notes, conducting research with AI assistance, engaging in multi-threaded conversations, and utilizing AI models for specific tasks without internet dependency.

How to use

To use MAI, clone the repository, install the necessary dependencies, and run the application either for web or Electron development. Users can select and run AI models locally on their devices.

Key features

Key features include local model inference, privacy by design, a rich chat interface, note-taking system, knowledge management with RAG capabilities, MCP integration, search capabilities, theme support, and fully offline functionality.

Where to use

MAI can be used in various fields such as personal productivity, education, research, and any domain requiring local AI assistance without compromising data privacy.

MAI - Model-Assisted Intelligence

A powerful desktop chat application that runs AI models locally, built with Electron, SvelteKit, and WebLLM.

Select and run any model from MLC.AI Models locally on your device.

Overview

MAI is a desktop application designed to provide AI assistance while respecting your privacy. All AI processing happens locally on your machine without sending your data to external servers.

DISCLAIMER: This application is under active development and is NOT intended for production use. It is provided “as is” without warranties.

Key Features

• Local Model Inference - Run AI models directly on your device using WebLLM
• Private by Design - No data leaves your computer unless explicitly requested
• Rich Chat Interface - Modern UI with support for multiple conversation threads
• Note-Taking System - Create, edit, and organize notes with AI assistance
• Knowledge Management - RAG (Retrieval-Augmented Generation) capabilities for your notes
• MCP Integration - Model Context Protocol support for advanced tool usage
• Search Capabilities - Find information online as context for the LLM
• Theme Support - Switch between dark and light modes
• Fully Offline - Works without an internet connection

Installation

Prerequisites

• Node.js (v18 or later)
• npm or pnpm

Development Setup

1. Clone the repository:

   git clone <repository-url>
   cd mai
   

2. Install dependencies:

   npm install
   # or
   pnpm install
   

3. Start the development server:

   # For web development
   npm run dev
   
   # For Electron development
   npm run electron-dev
   

Building for Distribution

# Build the application
npm run build

# Package as an Electron app
npm run pack:electron

Architecture

MAI combines several technologies to create a seamless local AI experience:

• SvelteKit - Frontend framework with file-based routing
• Electron - Desktop application wrapper with native capabilities
• WebLLM - WebAssembly-based LLM inference for running models locally
• IndexedDB - Client-side storage for chats, notes, and vector embeddings
• TypeScript - Type safety throughout the codebase
• TailwindCSS - Utility-first CSS framework for styling

Technical Details

• WebAssembly and WebGL acceleration for optimal model performance
• Context isolation in Electron for enhanced security
• Custom CSP headers configured for model loading
• Offline-first architecture with local storage

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT