LLM Wiki is an open-source desktop application on GitHub, with more than 20,800 stars, that turns your documents into an organized, interlinked knowledge base โ automatically. Based on Andrej Karpathy’s LLM Wiki pattern, it takes a different approach from conventional retrieval-augmented generation: instead of re-deriving answers from raw chunks on every query, an LLM incrementally builds and maintains a persistent wiki from your sources, so knowledge is compiled once and kept current. The project is actively developed, with updates landing in October 2026.
Built with a Tauri and Rust backend, LLM Wiki ingests PDFs, Office documents, ebooks, images, and web clips, then generates structured wiki pages with cross-references, source traceability, and YAML frontmatter. A four-signal knowledge graph with Louvain community detection surfaces surprising connections and knowledge gaps, while a built-in chat agent answers questions grounded in your material. The resulting wiki folder stays compatible with Obsidian, so your knowledge base remains portable plain Markdown.
LLM Wiki is free and open source. You bring your own LLM provider keys and pay only for the model usage you consume.
LLM Wiki suits researchers, writers, and knowledge workers who want a personal knowledge base that organizes itself, and anyone who prefers a persistent, inspectable wiki over opaque chat-over-documents tools. Because everything is stored as local Markdown files and you choose the model provider, it also fits privacy-conscious users who want AI-maintained notes without handing their library to a hosted service.