Privacy-safe AI tool that anyone can use without knowing the AI powered by the Local LLM technology.
- ChatGPT-like conversational interface with multiple LLM models
- Auto summarization of the conversation
- Multiple OS support (Windows, macOS, Linux)
- Easy-to-use GGUF based models
- Packaging: Electron
- Frontend: Vite, React, Tailwind CSS, shadcn/ui
- LLM(GGUF) inference: llamafile
- Electron packaging
- Manage(Spawn/Kill) local LLM models using llamafile
- (Frontend) Conversation UI
- Auto summarization of the conversation
- Tune the summarization prompt
- Using multiple LLM models in single conversation
- Using OpenAI API with local models
- Download GGUF models from the Huggingface
- Chat with uploaded files in the conversation (RAG)
- Upload files to the conversation
- Manage Vector DB