AI That Lives in Your Browser - A private, secure, and powerful AI chat interface that runs entirely in your web browser using WebLLM and WebGPU.

🚀 Try it here

✨ Features

• 100% Local Processing - No data ever leaves your browser
• Multiple AI Models - Choose from 5 different models
• Real-time Chat - Fast, responsive conversations with markdown support
• Conversation History - Automatic saving and export options
• Modern Interface - Dark theme, responsive design

🚀 Getting Started

Prerequisites

• Chrome 113+ or Edge 113+ (WebGPU support required)
• 8GB+ RAM and discrete GPU recommended

Installation

1. Clone this repository
2. Open index.html in your browser
3. Wait for model loading (~2.4GB download on first run)
4. Start chatting!

git clone https://github.com/your-username/browsermind.git
cd browsermind
python -m http.server 8000  # Optional local server

🤖 Available Models

Model	Size	Best For
Phi-3.5 Mini	~2.4GB	General chat, coding, reasoning
Llama 3.2 3B	~2.0GB	Strong reasoning, coding
Qwen 2.5 3B	~2.1GB	Multilingual, coding, chat
Gemma 2 2B	~1.4GB	Fast responses, efficiency
Llama 3.2 1B	~0.9GB	Basic chat, minimal resources

🔒 Privacy & Security

• No external servers - Everything runs locally
• No telemetry - No usage data collected
• Local storage only - Conversations saved in browser
• Sandboxed execution - Browser security model

📄 License

MIT License - see LICENSE for details.