"Panta rhei" - "Everything is in flow" - Herakles
A revolutionary web-based operating system interface powered by artificial intelligence that can create and modify applications on demand using natural language commands and voice input.
- Clone the Repository:
git clone https://github.com/CarstenMaul/ai-os.git cd ai-os - Run Locally:
python -m http.server 8000 --bind 0.0.0.0 - Open:
http://127.0.0.1:8000/ai-os.html - Setup: Complete the Setup Assistant or add your API keys in Settings.
- Create: Say "Create a calculator app" or type your request.
Your apps and data will automatically persist between sessions!
For the best experience and to avoid potential browser security restrictions (CORS), run ai-os.html from a local webserver.
Navigate to the project directory in your terminal and run:
# For Python 3
python -m http.server 8000 --bind 0.0.0.0- Node.js:
npm install -g http-server && http-server -p 8000 -a 0.0.0.0 - VS Code: Use the "Live Server" extension.
- AI-Powered App Generation: Create functional applications from natural language or screenshots.
- Real-time App Modification: Modify apps on-the-fly using voice or text.
- Advanced Version Control: Full snapshot management with version history.
- 3D & Visualization: Built-in support for Three.js and Chart.js.
- API Integration: Connect external APIs to supercharge your generated apps.
- Persistent Storage: Your apps and data are automatically saved locally.
AI-OS requires API keys for its core functionality. You can configure these in the Settings (⚙️) panel after launching.
- LLM Provider (Choose one):
- OpenRouter (Recommended)
- OpenAI
- LMStudio (for local models)
- OpenAI API Key (for Whisper voice recognition).
- Simple/Fast:
google/gemini-2.5-flash-lite-preview-06-17 - Tough/Advanced:
anthropic/claude-sonnet-4
For a deep dive into all features, architecture, development guidelines, and more, please see our comprehensive technical documentation.
➡️ Read the Full Documentation
This project is licensed under the terms of the LICENSE file.