Anything LLM
AnythingLLM is an all-in-one AI application that lets you build a private, fully-featured ChatGPT. It connects to your preferred LLM, ingests documents, and provides built-in agents, multi-user support, and vector databases with no extra configuration. It is for developers and teams who want a self-hosted, customizable AI chat solution.
![[FREE] AnythingLLM v2 | The last document chatbot you will ever need](https://i.ytimg.com/vi/f95rGD9trL0/hqdefault.jpg)
🎬 [FREE] AnythingLLM v2 | The last document chatbot you will ever need · Tim Carambat
✨ Key features
- Dynamic model routing to best provider per conversation.
- Automatic and user-managed memories for context.
- Scheduled tasks with full agent capabilities.
- Intelligent skill selection reduces token usage up to 80%.
- No-code AI agent builder and MCP compatibility.
- Multi-user support with permissioning (Docker only).
🎯 Use cases
- Build a private ChatGPT-like assistant for your team.
- Chat with your documents (PDF, DOCX, etc.) with citations.
- Automate workflows with AI agents that browse the web.
- Deploy a multi-user AI platform with controlled access.
- Embed a chat widget on your website (Docker only).
📦 Installation
🧰 Requirements: Requires Node.js and yarn for development; Docker for multi-user deployment. Supports many LLM providers, embedders, and vector databases; some require API keys.
For development:
yarn setupyarn dev:server
yarn dev:frontend
yarn dev:collectorFor self-hosting, refer to the README for Docker and cloud deployment options.
🚀 Usage
After setup, run the development servers:
yarn dev:server
yarn dev:frontend
yarn dev:collector
Then access the frontend at the provided URL and start chatting with your documents.
⚠️ Good to know
Multi-user support and custom embeddable chat widget are only available in the Docker version. Telemetry is on by default but can be disabled.
❓ FAQ
How do I opt out of telemetry?
Set DISABLE_TELEMETRY to 'true' in your server or docker .env settings, or disable it in-app via the sidebar > Privacy.
What LLMs are supported?
AnythingLLM supports many providers including OpenAI, Anthropic, Google Gemini, Ollama, and any llama.cpp compatible model.
Can I use it with multiple users?
Yes, multi-user support is available in the Docker version, with permissioning and access control.
What vector databases are supported?
It supports LanceDB (default), PGVector, Pinecone, Chroma, Weaviate, Qdrant, Milvus, and more.
📊 Repository
🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.