Horizon AI
🚀 Your Private Local AI Assistant
A beautiful desktop application for running AI models locally on your computer. 100% private. No cloud. No data sent to the internet.
Install • Features • Screenshots • Architecture • Development
⬇️ Quick Install
Windows (Recommended)
Download the installer and run it. That’s all!
📥 Download Horizon AI Installer (Windows x64)
The installer will automatically:
- Install Horizon AI on your computer
- Install Ollama (AI engine) silently if not present
- Download a default AI model (~2GB)
- Configure everything automatically
**System Requirements…
Horizon AI
🚀 Your Private Local AI Assistant
A beautiful desktop application for running AI models locally on your computer. 100% private. No cloud. No data sent to the internet.
Install • Features • Screenshots • Architecture • Development
⬇️ Quick Install
Windows (Recommended)
Download the installer and run it. That’s all!
📥 Download Horizon AI Installer (Windows x64)
The installer will automatically:
- Install Horizon AI on your computer
- Install Ollama (AI engine) silently if not present
- Download a default AI model (~2GB)
- Configure everything automatically
System Requirements:
- Windows 10/11 (64-bit)
- 8 GB RAM minimum (16 GB recommended)
- 10 GB free disk space
- NVIDIA GPU recommended (but not required)
✨ Features
🤖 Local AI Chat
- Chat with AI models running 100% on your computer
- Real-time streaming responses
- Conversation history with automatic saving
- Multiple AI models support
🔒 Complete Privacy
- No internet required after initial setup
- No data sent to external servers
- All conversations stored locally
- Full control over your data
🎨 Beautiful Interface
- Modern, clean design
- Dark/Light mode
- Interactive onboarding tour
- Responsive layout
⚡ Smart Features
- Automatic Ollama management (start/stop with app)
- GPU acceleration (NVIDIA CUDA)
- System monitoring (CPU, RAM, VRAM)
- Model management (install/delete models)
📸 Screenshots
Dashboard
The main dashboard with quick chat, system status, and model recommendations.
AI Chat
Full-featured AI chat interface with conversation history and code-friendly responses.
Settings
Application configuration panel for language, storage, and system preferences.
🏗️ Architecture
Horizon AI is built with a modern hybrid architecture:
┌─────────────────────────────────────────────────────┐
│ HORIZON AI │
├─────────────────────────────────────────────────────┤
│ Frontend (React + Vite) │
│ ├── Dashboard │
│ ├── AI Chat Panel │
│ ├── Settings │
│ └── File Manager │
├─────────────────────────────────────────────────────┤
│ Desktop Shell (Tauri / Rust) │
│ ├── Window Management │
│ ├── Ollama Installer │
│ ├── Python Bridge (IPC) │
│ └── System Integration │
├─────────────────────────────────────────────────────┤
│ Backend Worker (Python) │
│ ├── Chat History Service │
│ ├── Ollama Service │
│ ├── GPU Monitoring │
│ └── System Services │
├─────────────────────────────────────────────────────┤
│ AI Engine (Ollama) │
│ ├── LLaMA 3.2 │
│ ├── Mistral │
│ ├── DeepSeek │
│ └── + Many more models │
└─────────────────────────────────────────────────────┘
Tech Stack
| Component | Technology |
|---|---|
| Frontend | React 18, Vite, TailwindCSS |
| Desktop Shell | Tauri 2.0 (Rust) |
| Backend Worker | Python 3.11 (PyInstaller) |
| AI Engine | Ollama |
| IPC | JSON over stdin/stdout |
🧠 Supported AI Models
| Model | Size | Best For |
|---|---|---|
| llama3.2:3b | 2 GB | General use, fast responses |
| mistral | 4 GB | Balanced, versatile |
| deepseek-r1:7b | 4 GB | Reasoning, coding |
| codellama | 4 GB | Programming |
| phi | 2 GB | Lightweight, fast |
| gemma2 | 5 GB | Quality responses |
Install more models directly from the app dashboard!
💻 Development
Prerequisites
- Node.js 18+
- Rust (latest stable)
- Python 3.11+
- Ollama installed
Setup
# Clone the repository
git clone https://github.com/GabrielHori/Horizon-AI.git
cd Horizon-AI
# Install frontend dependencies
npm install
# Create Python virtual environment
python -m venv .venv
.venv\Scripts\activate # Windows
# source .venv/bin/activate # Linux/Mac
# Install Python dependencies
pip install -r worker/requirements.txt
# Run in development mode
npm run tauri dev
Build for Production
# Build Python backend
pip install pyinstaller
python -m PyInstaller --clean worker/backend.spec
copy dist\backend.exe src-tauri\binaries\backend-x86_64-pc-windows-msvc.exe
# Build Tauri application
npm run tauri build
The installers will be generated in:
src-tauri/target/release/bundle/nsis/Horizon AI_1.0.0_x64-setup.exesrc-tauri/target/release/bundle/msi/Horizon AI_1.0.0_x64_en-US.msi
📁 Project Structure
horizon-ai/
├── src/ # React frontend
│ ├── components/ # UI components
│ ├── pages/ # Page components
│ ├── services/ # API services
│ └── contexts/ # React contexts
├── src-tauri/ # Tauri (Rust) backend
│ ├── src/
│ │ ├── lib.rs # Main app logic
│ │ ├── python_bridge.rs # Python IPC
│ │ └── ollama_installer.rs # Ollama management
│ └── binaries/ # Sidecar binaries
├── worker/ # Python backend
│ ├── main.py # Entry point
│ ├── services/ # Business logic
│ └── ipc/ # IPC handlers
└── data/ # User data (conversations)
🔧 Configuration
Environment Variables
| Variable | Description | Default |
|---|---|---|
| OLLAMA_HOST | Ollama server URL | http://localhost:11434 |
| OLLAMA_MODELS | Models storage path | System default |
Settings (in-app)
- Language: French / English
- User Name: Your display name
- Storage Path: Custom Ollama models location
- Internet Access: Enable/disable network features
🐛 Troubleshooting
App doesn’t start
- Make sure you have enough disk space (10 GB+)
- Try running as Administrator
- Check if antivirus is blocking the app
Chat doesn’t respond
- Check if a model is selected (top bar)
- Verify Ollama is running (green status indicator)
- Try selecting a different model
Slow responses
- Close other resource-heavy applications
- Use a smaller model (llama3.2:3b, phi)
- If you have an NVIDIA GPU, it will be used automatically
📄 License
MIT License - See LICENSE file for details.
🙏 Credits
Made with ❤️ by Gabriel Hori
Horizon AI v1.0 • Powered by Local AI ⚡