AnythingLLM
AnythingLLM is a self-hosted, multi-user application for chatting with a user's own documents and building custom AI agents, connecting to almost any local or cloud LLM, embedder and vector database.
Features include dynamic routing of a chat to the best configured provider and model, automatic and user-managed long-term memory, scheduled agent tasks on a cron schedule, a no-code agent builder, MCP compatibility, multimodal support for both closed and open-source models, and a custom embeddable chat widget in the Docker build. It ingests multiple document types including PDF, TXT and DOCX, and supports per-user access control in multi-user deployments.
Supported LLM providers include OpenAI, Azure OpenAI, AWS Bedrock, Anthropic, Google Gemini, Ollama, LM Studio, DeepSeek, Mistral, Groq and several others, with a matching list of embedder and vector-database options. Available as a desktop app for Mac, Windows and Linux, and as a separate open-source Android app.
JavaScript, MIT licensed.