Skip to main content

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.

AnythingLLM is designed to go from zero to a working private AI assistant as fast as possible. The quickest path is Docker — a single command pulls the image, mounts a local storage folder, and starts the server. From there, a guided setup wizard walks you through every remaining step in the browser. If you’d rather skip Docker entirely, the desktop app gets you running just as quickly with no command-line work at all.
Prefer a GUI installer? Download the AnythingLLM Desktop App for Mac, Windows, or Linux. It bundles everything — including local model support — and requires no Docker or terminal commands. Perfect for non-technical users or anyone who wants a purely local setup.
If you plan to run AnythingLLM on bare metal (without Docker), you need Node.js v18 or higher and Yarn installed before you begin. See the Bare Metal installation guide for the full walkthrough.

​
Prerequisites

  • Docker installed and running
  • Access to an LLM — either a local model via Ollama or an API key from a provider like OpenAI
1

Pull and run the Docker image

Run the following command in your terminal. It pulls the latest AnythingLLM image, creates a persistent storage folder in your home directory, and starts the server on port 3001.Linux / macOS:
export STORAGE_LOCATION=$HOME/anythingllm && \
mkdir -p $STORAGE_LOCATION && \
touch "$STORAGE_LOCATION/.env" && \
docker run -d --rm -p 3001:3001 \
--cap-add SYS_ADMIN \
-v ${STORAGE_LOCATION}:/app/server/storage \
-v ${STORAGE_LOCATION}/.env:/app/server/.env \
-e STORAGE_DIR="/app/server/storage" \
mintplexlabs/anythingllm
Windows (PowerShell):
$env:STORAGE_LOCATION="$HOME\Documents\anythingllm"; `
If(!(Test-Path $env:STORAGE_LOCATION)) {New-Item $env:STORAGE_LOCATION -ItemType Directory}; `
If(!(Test-Path "$env:STORAGE_LOCATION\.env")) {New-Item "$env:STORAGE_LOCATION\.env" -ItemType File}; `
docker run -d --rm -p 3001:3001 `
--cap-add SYS_ADMIN `
-v "$env:STORAGE_LOCATION`:/app/server/storage" `
-v "$env:STORAGE_LOCATION\.env:/app/server/.env" `
-e STORAGE_DIR="/app/server/storage" `
mintplexlabs/anythingllm;
The image supports both amd64 and arm64 CPU architectures. The -v volume mounts ensure your data persists across container restarts and upgrades.
If you’re connecting AnythingLLM to a service running on your host machine (such as Ollama on localhost:11434), use http://host.docker.internal:11434 instead of localhost in your configuration. On Linux, add --add-host=host.docker.internal:host-gateway to the docker run command.
2

Open the setup wizard

Once the container is running, open your browser and navigate to:
http://localhost:3001
The first time you visit, AnythingLLM presents a setup wizard. Follow the prompts to:
  1. Create your admin account (username and password)
  2. Choose your LLM provider (covered in the next step)
  3. Choose your embedding engine (the native built-in embedder works out of the box)
  4. Choose your vector database (LanceDB is the default and requires no configuration)
You can change every setting later from the admin panel, so don’t worry about getting everything perfect right now.
3

Configure an LLM provider

AnythingLLM supports 30+ LLM providers. Choose the option that fits your setup:
# 1. Make sure Ollama is running on your host machine:
#    https://ollama.ai/download
#
# 2. Pull a model (e.g. llama3.2):
ollama pull llama3.2
#
# 3. In the AnythingLLM setup wizard, select "Ollama" as the LLM provider
#    and enter the base URL:
#    http://host.docker.internal:11434
#
# 4. Select your pulled model from the dropdown (e.g. llama3.2)
After saving your LLM settings, AnythingLLM will verify the connection. A green checkmark confirms the provider is reachable and the model is available.
4

Create a workspace and upload a document

Workspaces are isolated chat environments — each one has its own document set, system prompt, and settings. Think of them as separate AI assistants for different projects or topics.
  1. From the home screen, click “New Workspace” and give it a name (e.g. My First Workspace).
  2. Once inside the workspace, click the document upload icon (paperclip or upload button in the sidebar).
  3. Drag and drop a PDF, DOCX, TXT, or other supported file. AnythingLLM will automatically chunk, embed, and index it.
  4. Toggle the document on in the workspace panel to make it available for RAG retrieval.
Supported document types include PDF, TXT, DOCX, CSV, JSON, XML, and many more. You can also paste raw text, link to web pages, or connect data sources directly from the UI.
5

Send your first chat and explore agents

You’re ready to chat. Type a question in the message box — AnythingLLM will search your uploaded documents and respond with grounded answers, showing source citations below the response.Enable AI Agents (optional):Agents can browse the web, execute code, summarize files, and more. To enable them:
  1. Open Workspace Settings (the gear icon next to the workspace name).
  2. Navigate to the Agent tab and toggle on the skills you want (e.g. web browsing, code execution).
  3. Switch the chat mode from Chat to Agent using the selector in the message bar.
  4. Send a message like Search the web for the latest news about AI agents to see it in action.
You can also build no-code agent workflows using the Agent Flows builder in the admin panel. Agent flows let you chain tools, conditions, and LLM calls into repeatable automations.

​
Next Steps

Docker Installation

Learn about volume mounts, environment variables, GPU support, and Docker Compose.

Desktop App

Install the native desktop app for a zero-setup local AI experience.

Bare Metal Setup

Run AnythingLLM directly on Linux or macOS without containers.

Configuration Reference

Full list of environment variables and advanced configuration options.

Build docs developers (and LLMs) love

Get started for freeTalk to us