Brethof AI Reviews

AnythingLLM Desktop review

Chat with your documents, no account, finds your Ollama

Not sponsoredModels & AI Infrastructure

Tested on Linux · Free

AnythingLLM is an all-in-one app for chatting with local or cloud AI models and your documents. We tested the Desktop 1.18.0 AppImage on Linux with a cloud model through our Ollama.

The good

  • No account, finds Ollama. No account at all. Pick Ollama and it finds your local Ollama by itself, fills in its address and lists its models. It also has its own engine.
  • PDF chat with citations. A PDF in a workspace embeds on your computer in seconds, and every answer shows its source. All three were right. For one, default Agent mode also searched You.com unasked.
  • About 40 providers. Besides its own engine, it lists about 40 providers, including Ollama, LM Studio, KoboldCPP and any OpenAI-compatible server.

The not so good

  • Desktop app is proprietary. Only the core and the Docker server are MIT open source. The desktop app is proprietary, for personal and internal business use only, no redistribution, no reverse engineering.
  • Telemetry on by default. Anonymous telemetry is on from the start, and the setup privacy page never mentions it. The first event leaves before any screen, with no chat content. The switch is in Settings, Privacy and Data.

Our verdict

Recommended with caveats

No account, it finds your Ollama, and document chat cites its sources. But default Agent mode sent a question to You.com unasked, telemetry is on before the first screen without a word in setup, and the desktop app is proprietary.

Reviewed: AnythingLLM Desktop 1.18.0 · facts checked 2026-10-21

Watch it onYouTube

Not sure which tool fits your setup? Tech advice: ask us before you spend yours. Describe your situation and get our answer in writing within two working days.

Ask us →

← All reviews

Hear it when it ships

New releases, real benchmarks and the occasional deep-dive. No spam, unsubscribe in one click.

Everything we build