KiloSnow Desktop AI Assistant

A simple desktop AI assistant that uses static character illustrations

Character

• Concision - Use static character illustrations
• Intelligent dialogue - AI conversation function based on Ollama
• Speech synthesis - Optional voice reading with Edge-TTS
• Transparent window - Borderless transparent window, beautify the desktop
• Right-click menu - Right-click to show exit menu

Screenshot

Environmental requirements

• Python 3.9 or a higher version
• Ollama (For AI conversations)
• Windows operating system (restricted by PyQt6)

Installation steps

1. Clone the repository.

git clone https://github.com/PDYAM/KiloSnow.git
cd KiloSnow

2. Install Python dependencies

pip install -r requirements.txt

3. Start the Ollama service.

ollama serve

4. Create or download a model.

ollama pull qwen3
# Or use other models, such as llama3:
# ollama pull llama3

5. Run the program

python src/main.py

Or simply double-click the start.bat file (Windows)

Instructions for use

Basic operations

• Right-click: Display the exit menu.
• Input box: Enter messages to chat with AI
• Send button: Send messages
• Voice button: Enable/Disable voice reading

Configuration instructions

Edit the config.json file to customize settings:

{
  "ollama": {
    "model": "kiloSnow"  //Modify it to the name of the Ollama model you are using.
  },
  "tts": {
    "enabled": true,  //Whether to enable voice.
    "voice": "zh-CN-XiaoxiaoNeural"  //Voice type
  }
}
### Replace the character's illustration.



Place your character image in the `assets/` directory and name it `kiloSnow.jpg`, or modify the image path in `index.html`:


```html
<img src="assets/your-image.jpg" alt="Character" class="character">

Project structure

KiloSnow/
├── src/
│   └── main.py              # Main program
├── assets/
│   ├── kiloSnow.jpg         # Character illustrations
│   └── audio/               # The generated audio file
├── docs/
│   └── DEVELOPMENT.md       # Development documentation
├── index.html               # Front-end interface
├── config.json              # Configuration file
├── requirements.txt         # Python dependencies
├── .gitignore              # Git ignore files
├── start.bat               # Windows startup script
├── README.md               # Project description
└── LICENSE                 # License

Technology stack

• Desktop framework: PyQt6 + QtWebEngine
• Web framework: Flask
• AI model: Qwen3
• Speech synthesis: Edge-TTS (Optional)
• Front-end: HTML + CSS + JavaScript

Contribution

Welcome to contribute code, report issues or submit suggestions!

1. Fork this repository
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Submit changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open Pull Request

Plan

• Add more animation effects
• Support voice recognition input
• Add more custom options

Feasible changes

1. Using Live2D models makes characters more three-dimensional

2. Change to other sound sources and use GPT-SoVITS or GLM-TTS, etc

3. Adding local memory function for characters

4. Manage your desktop realistically or observe your real desktop!

Common Questions

1. The window cannot be displayed.

• Check whether PyQt6 and PyQt6-WebEngine are installed
• Confirm that the Python version is >= 3.9

2. The AI is unresponsive.

• Make sure the Ollama service (ollama serve) is running
• Check if the Ollama model has been downloaded
• Check whether the configuration in config.json is correct

3. The voice function is not working.

• Make sure that edge-tts is installed (pip install edge-tts).
• Check if the voice function is enabled (click the "Voice" button)

License

This project is open source under the MIT License.

Acknowledgements

• Ollama - AI model service
• PyQt6 - Python GUI framework
• Edge-TTS - Microsoft Edge TTS

Contact information

• Project homepage: https://github.com/PDYAM/KiloSnow
• Problem feedback: GitHub Issues

---

**If you find this project helpful, please give a star**