Ask your AI what a photo reveals

Connect a model running on your own machine — Ollama or LM Studio — and get a privacy read of any photo: faces, documents, reflections, location clues. The photo travels from this page to the model server on your computer and nowhere else. We never see it, and there is nothing to pay.

Connect your model server

1 · Allow this site to talk to it

Ollama only answers websites you explicitly allow. Run the line for your OS once, then restart Ollama:

macOS

launchctl setenv OLLAMA_ORIGINS "https://metamarshal.com"
# then quit & reopen the Ollama app

Linux

sudo systemctl edit ollama
# add under [Service]:
Environment="OLLAMA_ORIGINS=https://metamarshal.com"
sudo systemctl restart ollama

Windows

setx OLLAMA_ORIGINS "https://metamarshal.com"
REM then quit Ollama from the tray and reopen it

2 · Find it

Chrome may ask permission to reach your local network the first time — that prompt is the browser doing its job; allow it.

Why this exists: cloud AI tools can tell you a lot about a photo — which means uploading your photo to someone else's computer. If you already run Ollama or LM Studio, you get the same insight with zero disclosure. Works in Chrome and Firefox; Safari blocks pages from reaching localhost. Don't have a local model? The reveal still shows everything the metadata says without any AI at all.