nodx Lens puts an AI action wheel over everything on your phone — and it takes
your own instructions.
Tap the floating bubble, drag over any region of any app — a product photo, a
paragraph, a chart — and pick a spoke:
✏️ Instruct — type what you want done with what you boxed. Plain words work:
ask a question and the answer streams into a card you can keep talking to; say
"find it on Amazon" or "search this on arXiv" and Lens routes the intent and
opens the real results page.
💬 Explain — vision AI answers about the exact pixels you boxed. Answers render
as readable formatted text, and every answer takes follow-up questions.
🔍 Search — AI names what it sees, then opens the destination you choose from a
built-in library of search and shopping sites.
💡 Save — the crop lands in your photo library.
🎨 Generate — AI describes the subject, then renders a single image or a 2×2
style grid.
EVERYTHING IS LOGGED
Prompts, full answers, instructions, search links, generated images — reopen
any of them from the activity log. The action wheel itself is fully
customizable: icons, names, colours, prompts, and submenus.
BRING YOUR OWN KEY
Choose your AI provider and paste your own key. A free-model option and a
no-key local option are both supported. Keys are stored on your device and
requests go directly to the provider — there is no nodx server, no account,
no subscription, no telemetry.
ONE-TIME CONSENT, SILENT CAPTURE
Grant accessibility access once; after that, capture works without repeated
screen-share dialogs.
OPEN SOURCE
Apache-2.0, code on GitHub (aistoume/nodx). nodx Lens is the mobile entry
point of nodx — a decision-thinking workspace where AI helps you think more,
not less.
Box any screen region → instruct AI to explain, search, shop, or generate.