Select text or box any region on any webpage — an action wheel appears: explain, search, shop, or generate images. Results stream in place, captures stay marked on the page as reusable action hubs. No backend, bring your own key, open source, free forever.

Hi Product Hunt! I built nodx Lens because my browsing had turned into a copy-paste relay: see something → copy → switch to a chat tab → paste → lose the answer forever. Lens flips it: select text or box any region of the page, and a four-spoke action wheel appears right there — Explain, Search, Shop, or Generate. The answer streams in place, and the capture stays marked on the page. Click the box next week and keep working on it. What's special: * v1.0 makes the wheel fully yours — every spoke's icon, action, prompt, and submenu is editable, with a live preview * Bring your own key: Anthropic, OpenAI, Google, or OpenRouter's free models ($0). No account, no subscription, no nodx server — everything lives in your browser * 100% local-first and open source (Apache-2.0) * An Android companion ships too — same wheel, system-wide via floating bubble It's free forever. I'll be here all day — tell me what's confusing, what's missing, and what you'd want the wheel to do next. 🙏

A custom prompt option would be huge, letting me type my own instruction for the selected text instead of only the four preset actions. Would make it way more useful for niche workflows I run daily.

@dnebhqs Thanks for your trying. It is a wonderful suggestion and a real use case for us. Adding them in the coming version.

A custom prompt option would be huge, letting me type my own instruction for the selected text instead of only the four preset actions. Would make it way more useful for niche workflows I run daily.

@dnebhqs Thanks for your trying. It is a wonderful suggestion and a real use case for us. Adding them in the coming version.

nodx Lens

Turn anything you see on a page into AI action

Turn anything you see on a page into AI action