See exactly how ChatGPT picks its sources.

Yes. There's no paid tier or trial counting down. I built it because I wanted it to exist.

Both work and both are the same code. The Chrome Web Store is one click and it updates itself, so start there. The catch is that ChatGPT changes its internal format often, sometimes twice in a week, and every store update waits in Google's review queue for a day or three. When a format change breaks captures I can put the fix on the download page the same day. The store copy catches up later. Keep that page bookmarked for those days.

No. It all runs on your machine. Captures sit in your browser's own storage and are never uploaded. Two optional lookups make the only outbound calls, Ahrefs domain ratings and Reddit thread stats, and you can switch both off.

ChatGPT and Perplexity. Perplexity reads the same way, the fan-out, the sources and the citations, and it also returns per-source trust notes that nothing else surfaces. The Compare view puts the same prompt side by side across both engines. Gemini is next. One capture format covers all of them, so your saved sessions carry over.

Generative Engine Optimisation. It's the work of getting your brand picked, cited and mentioned inside AI answers instead of ranked on a results page. It's what SEO becomes when the answer replaces the results page.

No. On ChatGPT it uses the session token your browser already holds, once, to fetch your own conversation, then throws it away. It's never stored or sent anywhere. On Perplexity it reads no token at all, because it only watches the answer your browser is already receiving.

No. It does nothing at all until you click the fox. It never intercepts, proxies or changes your conversations.

Yes. Capture the answers for a client's queries, pin them, run them again later and compare who entered or dropped out. Plenty of people use it as a light GEO rank tracker across several client brands.