AnythingLLM Download: Chat With Your Documents Using Local AI
AnythingLLM · for Windows, macOS and Linux
Developer: Mintplex Labs (opens in a new tab)
- Windows
- macOS
- Linux
The download link points to the program’s official source; the file is not hosted on this site. Link address: anythingllm.com
Link not working? Let us know
AnythingLLM is a free, open-source all-in-one AI app for chatting with your documents, building agents and using local or cloud models in one workspace. The desktop app runs on Windows, macOS and Linux, a Docker version serves teams, and it is MIT-licensed.
AnythingLLM, by Mintplex Labs, is built around one common wish: “let me ask questions about my own files.” You drop in documents, pick a model and chat, with answers that cite their sources. By default it runs locally, with its own built-in model engine.
What does AnythingLLM do?
- Chat with documents: Drag and drop files and ask questions; answers include source citations.
- Local or cloud models: A built-in local LLM engine (CPU, NVIDIA via CUDA, AMD and Intel via Vulkan), plus Ollama, LM Studio, OpenAI, Anthropic, Google Gemini, Mistral, Groq and many others.
- Vector databases: LanceDB by default, with several other options.
- Agents: Built-in agents and a no-code agent builder, with Model Context Protocol (MCP) support.
- Other features: Scheduled tasks, dynamic model routing and memories. Multi-user permissions are available in the Docker version.
Who is it for?
It suits individuals and small teams who want a private knowledge assistant without stitching several tools together.
Getting the desktop app
The Download button opens the official download page; pick the desktop app for Windows, macOS or Linux and install it like any other program. A Docker version for servers and multi-user setups is documented in the project repository.
AnythingLLM collects anonymous usage telemetry by default. You can switch it off in the app under Privacy in the sidebar, or with the DISABLE_TELEMETRY setting in the Docker version.
Recommended hardware
For a basic experience the documentation recommends:
- RAM: 16 GB
- CPU: An 8-core processor of any brand
- Storage: Depends mainly on the size of the local models you keep; with cloud models the requirements are much lower.
- Windows: A GPU with 8–12 GB or more of VRAM is recommended for faster local models.
- macOS: Any M-series chip can run local models; Intel Macs work but are slower.
AnythingLLM is free and open source under the MIT license.
Pros
- Chat with your documents, with cited sources
- Works with local and cloud models
- Agents and MCP support built in
Cons
- Telemetry is on until you turn it off
- Multi-user features need the Docker version
Frequently Asked Questions
Can AnythingLLM run without the internet?
Yes, if you use a local model such as its built-in engine or Ollama. Your documents and chats then stay on your computer.
What computer do I need for AnythingLLM?
The project recommends 16 GB of RAM and an 8-core CPU for a basic experience with local models. On Windows, a GPU with 8–12 GB of VRAM helps local models run faster; on a Mac, any M-series chip works.
AnythingLLMOpen source
Download (opens in a new tab)
