AnythingLLM

AnythingLLM is an all-in-one open-source application that transforms any document into an intelligent knowledge base, queryable through an AI chatbot. Compatible with dozens of LLMs (OpenAI, Anthropic, Ollama, LM Studio...), it offers complete flexibility in model and deployment choice.

The flagship feature is RAG (Retrieval Augmented Generation): import your PDFs, Word documents, web pages, or text files, and ask questions in natural language. AnythingLLM automatically retrieves relevant passages and generates sourced answers. Built-in AI agents can execute tasks like web browsing, chart generation, or data interaction.

The application is available in three modes: free desktop (Windows, Mac, Linux), self-hosting via Docker (free, 100% private), or hosted cloud (from $50/month). Multi-user mode with role and permission management makes it suitable for teams.

AnythingLLM stands out with its privacy-first philosophy: your data stays local if you choose, never sent to third-party servers. It's the ideal solution for privacy-conscious organizations wanting to leverage LLM power on their internal documents.