AnythingLLM is the all-in-one AI platform that puts you in control. Chat with your documents, automate workflows with AI agents, connect any LLM provider, and deploy privately—on your own hardware or in the cloud. No vendor lock-in, no data leaving your infrastructure unless you choose it.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/Mintplex-Labs/anything-llm/llms.txt
Use this file to discover all available pages before exploring further.
Quickstart
Get AnythingLLM running and send your first chat in under 5 minutes.
Docker Install
The recommended way to run AnythingLLM in production with a single command.
LLM Providers
Connect OpenAI, Anthropic, Ollama, Gemini, or any of 30+ supported providers.
API Reference
Integrate AnythingLLM into your applications with the full developer API.
What is AnythingLLM?
AnythingLLM is a production-ready, self-hosted AI application. You bring your documents and your preferred LLM; AnythingLLM handles everything else—ingestion, chunking, embedding, vector storage, retrieval, and a polished chat UI.RAG Chat
Upload PDFs, Word docs, URLs, and more. AnythingLLM indexes them and cites sources in every response.
AI Agents
Built-in agents browse the web, run SQL queries, read your filesystem, and execute custom workflows.
Multi-User
Role-based access control for teams—admins, managers, and members each get the right permissions.
30+ LLMs
OpenAI, Anthropic, Ollama, Gemini, Mistral, Groq, and many more—swap providers without re-indexing.
Vector DBs
Built-in LanceDB or connect Chroma, Pinecone, Qdrant, Weaviate, pgvector, and more.
Embed Widget
Drop a chat widget onto any website—scoped to a specific workspace, with full domain controls.
Get started
Install AnythingLLM
Run AnythingLLM via Docker (recommended), the desktop app, or directly on bare metal. See the installation guides for your platform.
Connect an LLM provider
Point AnythingLLM at your preferred LLM—from a local Ollama instance to OpenAI or Anthropic in the cloud. Configure your LLM provider.
Create a workspace and upload documents
Workspaces are isolated chat environments. Create one, upload your documents, and AnythingLLM handles chunking, embedding, and indexing automatically. Learn about workspaces.
Start chatting
Open the workspace chat and ask questions about your documents. Optionally enable AI agents to give the model tools like web search, code execution, and more.
Deployment options
Docker
Single-container deploy with
docker run. Recommended for servers and cloud VMs.Desktop App
Native app for Mac, Windows, and Linux. Zero config required—batteries included.
Bare Metal
Run Node.js processes directly without Docker for full control over the environment.
Cloud Platforms
One-click deploys on AWS, GCP, DigitalOcean, Railway, Render, and more.