RunAnywhere AI - On-Device AI Assistant

A privacy-first, offline-capable AI assistant that runs entirely in your browser. No cloud dependencies, no data collection, complete user control.

✨ Features

🔒 Privacy First

• 100% On-Device Processing - All AI runs locally in your browser
• Zero Cloud Dependency - Works completely offline after initial model download
• Privacy Mode - Toggle to prevent any cloud logging
• Full Data Control - Export or delete your data anytime

⚡ Performance

• Zero Network Latency - No API calls, instant responses
• WebGPU Acceleration - Hardware-accelerated inference
• Optimized Models - Fast, efficient on-device models
• Sub-100ms Response - First token in under 100ms

🎯 Capabilities

• 💬 Chat - Conversational AI assistant
• 📷 Vision - Image understanding and description
• 🎙️ Voice - Speech-to-text and text-to-speech

📊 Data Management

• Local Storage - All data stored in your browser
• CSV Export - Download your data as CSV
• JSON Export - Export in JSON format
• Full Privacy - No cloud, no external servers

🚀 Quick Start

Prerequisites

• Node.js 18+ and npm
• Modern browser with WebGPU support (Chrome 113+, Edge 113+)

Installation

# Clone the repository
git clone https://github.com/yourusername/runanywhere-ai.git
cd runanywhere-ai/web-starter-app

# Install dependencies
npm install

# Start development server
npm run dev

Visit http://localhost:5173 in your browser.

📦 Project Structure

web-starter-app/
├── src/
│   ├── components/       # React components
│   │   ├── ChatTab.tsx
│   │   ├── VisionTab.tsx
│   │   ├── VoiceTab.tsx
│   │   └── SettingsModal.tsx
│   ├── hooks/           # Custom React hooks
│   ├── workers/         # Web Workers for AI processing
│   ├── styles/          # CSS styles
│   ├── App.tsx          # Main app component
│   ├── runanywhere.ts   # SDK initialization
│   ├── privacy.ts       # Privacy controls
│   ├── dataControl.ts   # Data management
│   └── supabse.ts       # Optional cloud storage
├── public/
│   ├── sw.js           # Service Worker for offline support
│   └── manifest.json   # PWA manifest

🔧 Configuration

All data is stored locally in your browser. No external configuration needed.

Privacy Mode

Privacy mode is enabled by default. All data stays on your device:

• No data sent to cloud
• All processing on-device
• Data stored locally only

Toggle via the 🔒 button in the header or Settings (⚙️).

🎨 Features in Detail

Chat Assistant

• Powered by LiquidAI LFM2 350M model
• Streaming responses for real-time feedback
• Conversation history with stats
• Quick prompt suggestions

Vision Understanding

• Camera integration for live analysis
• Single-shot and live mode
• Customizable prompts
• Fast inference with optimized models

Voice Assistant

• Voice Activity Detection (VAD)
• Speech-to-Text (Whisper Tiny)
• Text-to-Speech (Piper TTS)
• Complete voice pipeline

📱 Progressive Web App

Install as a standalone app:

1. Visit the app in Chrome/Edge
2. Click the install icon in the address bar
3. Use like a native app, works offline!

🛠️ Development

# Development server
npm run dev

# Build for production
npm run build

# Preview production build
npm run preview

# Type checking
npm run type-check

🌐 Browser Support

Browser	Version	WebGPU	Status
Chrome	113+	✅	✅ Full Support
Edge	113+	✅	✅ Full Support
Firefox	121+	🚧	⚠️ Experimental
Safari	18+	🚧	⚠️ Limited

📄 License

MIT License - see LICENSE file for details.

🤝 Contributing

Contributions welcome! Please read our contributing guidelines first.

🙏 Acknowledgments

• RunAnywhere SDK - On-device AI framework
• LiquidAI - LFM2 models
• Supabase - Optional cloud storage

📞 Support

• 📧 Email: support@example.com
• 💬 Discord: Join our community
• 🐛 Issues: GitHub Issues

---

Built with ❤️ for privacy and performance