Tools for
Conviviality

Home / Local AI / AnythingLLM

AnythingLLM

A desktop or Docker app for chatting with your documents using local or cloud models.

Instead of ChatGPT, Claude

M · ToolOpen sourceFreeWorks with local models
github.com/Mintplex-Labs/anything-llm
Screenshot of the AnythingLLM website

Screenshot from github.com/Mintplex-Labs/anything-llm, taken 2026-10-30, shown to help you recognise the tool. It belongs to its maker. Ask us to remove it.

Drop a stack of PDFs onto the desktop, ask a question, and watch an assistant answer from your own files without anything leaving the house, and the project's README sells it with a line about owning your intelligence and not renting it. AnythingLLM runs on Mac, Windows and Linux or in Docker, and it can use local models on your CPU or GPU or connect to cloud ones.

AnythingLLM scores a 5 on lock-in and self-direction, a 4 on footprint, longevity and delivery and a 3 on power.

The code is MIT licensed and Mintplex Labs, a company, makes it and also sells a hosted instance. The README describes anonymous telemetry through PostHog that is on by default, with an opt-out in the settings or an environment variable, and it says even then you will still see outbound connections to the providers you chose, to its model download host and to GitHub. Multi-user mode, white-labelling and the embeddable widget exist only in the Docker version, according to the page.

    What stands out

  • Chats with your documents on Mac, Windows, Linux or Docker, using local or cloud models
  • MIT licensed, from a company that also sells a hosted version
  • Anonymous telemetry is on by default and can be turned off
Visit AnythingLLM
Do you make AnythingLLM? Get the badge for your site

This badge reports a dated review. It is not a certification and it costs nothing. Link it to this page and leave the numbers alone: the image always shows the current score.

For dark backgrounds, use anythingllm-dark.svg in place of anythingllm-light.svg.

For comparison, we also reviewed ChatGPT (14/30) and Claude (15/30) as a benchmark.

Other tools for “chat with ai on my own machine”