anything-llm

Why we included this project

AnythingLLM is a desktop and server app that keeps your data on your own machine: you point it at your documents, it ingests and embeds them, and it pulls the relevant passages to ground answers from whatever LLM you connect, local or cloud. The retrieval pipeline ships as one piece, so you don't have to assemble separate components yourself. It also runs as a multi-user web service with shared workspaces and per-user permissions, and the agent mode goes beyond plain Q&A into tool use and scheduled tasks. If you want a private knowledge assistant that works with your own models, storage, and vector database, this is a strong place to start.