This is an application I'm developing, the idea is to create a small system around llama.cpp tools but without need to hook with the dll files.
- Model management
- HuggingFace Integration
- Llama.cpp GitHub Integration with releases management
- Llama-server terminal launching with easy arguments customization, Internal / External
- Simple chat interface for easy testing
- Hardware monitor
- Color themes
- Cleanup code and organize it better
- Add option to group files by model / quantization / architecture
- Improve chat organization with model tagging and ability to rename the chat
- Tools (MCP) compatibility
- RAG implementation
- Make compatible with Linux and Mac
Use cargo tauri build, it will output to backend\target\release.
https://github.com/fredconex/Llama-OS/releases
