Mozilla Readability is great when a page looks like an article. Plenty of pages do not.
Documentation sites, discussion feeds, product pages, dashboards, recipes, and hand-built layouts all break a one-method converter in different ways. Either it misses useful content or copies the whole mess.
Page to Markdown starts with Readability for article-shaped pages. If that does not return enough, it looks for a semantic main region. If that still falls short, it cleans the rendered page by removing navigation, cookie banners, hidden elements, overlays, scripts, and controls with no useful text.
Selection mode skips all of that. It converts the selected content and keeps links, lists, tables, images, and code where the page exposes them.
The extension works from the rendered DOM. It cannot reliably read text inside a canvas, image, video frame, closed shadow root, protected PDF, or cross-origin frame. It does not load the rest of an infinite feed or send a URL to a remote scraper. There is no OCR or AI summary hiding behind the button.
Conversion happens on the device. The result stays in memory until the user copies it or downloads a Markdown file.