Page Scanner

Capture a whole web page as a PDF whose text is still text, selectable and searchable. Also PNG and JPEG.

Page Scanner turns a web page into a PDF that is actually a document, not a picture of one. One click captures the whole page, top to bottom, in a single layout. The editor opens in a new tab where you crop exactly the region you want, then export it. Text stays selectable, vector graphics stay sharp at any zoom, and the crop you draw on screen is the crop you get in the file. HOW IT WORKS • Click the toolbar icon, or press the keyboard shortcut (Alt+Shift+P unless Chrome has given it to something else), on the page you want. • The whole page is laid out once at its full height, and both the preview and the PDF are taken from that single layout. That is why the crop lines up perfectly with the output. • Drag the frame edges to crop, drag inside it to move it, scroll or pinch to zoom. • Extract, pick the format, and the file downloads. A PNG can go to the clipboard instead, so it pastes straight into whatever you were doing. WHAT YOU GET • Text that is still text. The PDF comes out of Chrome's own print-to-PDF pipeline, the same code path as Print then Save as PDF, so you can select a sentence, search the document and copy out of it. A tool that stitches screenshots into one bitmap cannot: everything in that file is pixels. • Vector artwork that stays sharp at 400% zoom, because it never became pixels. • The crop you drew, exactly. The preview and the PDF come from one shared layout of the page, so the frame you dragged on screen is the region in the file. • No host permissions at all, so the extension has no standing access to any site. • The crop laid onto A4 or US Letter sheets, with a half-inch margin so it can actually be printed, and the sheet cuts drawn onto the capture while you choose. Or one long page at the exact size of your crop, if you want the capture rather than a document. Each choice tells you what it will produce: how many sheets, how much it shrinks the page, and what that leaves your body text at in points. • PNG and JPEG of the same crop when you want an image instead, with a quality control for JPEG and the output pixel size shown before you commit. • Full page height, including everything below the fold, without scrolling or stitching. • A raster fallback that still produces a usable PDF on pages where the vector capture cannot run. Its PDF does embed an image, so its text cannot be selected; the editor labels the capture "Raster" and says so before you export, and you always know which one you got. • Light and dark themes. PRIVACY Page Scanner contains no analytics and collects no data. Captures live in your own browser and are deleted automatically: anything older than 12 hours is swept the next time the extension starts. There are no host permissions, so the extension has no standing access to any site: it only ever touches the one tab you scan, at the moment you scan it. Out of the box it makes no network requests at all. LOCAL AI AGENTS (OPTIONAL, OFF BY DEFAULT) If you want it to, Page Scanner can let an AI agent running on your own computer take a capture and receive the file, over MCP. You turn this on by pairing: a command prints a port and a token, you paste both into the settings page. While it is on the extension talks to 127.0.0.1 on your own machine and to no other address, and the token is what stops any other program on the machine asking for a scan. Turning it off closes the connection. A NOTE ON THE DEBUGGER BANNER Producing a true vector PDF requires Chrome's own print-to-PDF pipeline, which extensions reach through the debugger API. While a scan runs, Chrome shows a "Page Scanner started debugging this browser" bar at the top of the window. It is Chrome's notice, not the extension's, it is up for as long as the scan takes — a few seconds on an ordinary page, longer on a very large one — and it cannot be suppressed. Nothing is being inspected or sent anywhere. KNOWN LIMITS • Chrome does not let any extension read chrome:// pages, the Web Store, or other extensions' pages. • If DevTools is open on the tab, the vector capture cannot attach and the raster fallback is used instead. • Viewport-relative CSS (a 100vh hero, for example) is measured at the page's own viewport and held there, so a full-height hero does not stretch over the capture. A page that marks such a rule !important in its own inline style can still win. • Text a page paints into a <canvas>, as some map, chart and document viewers do, is not text to the browser either, so it arrives as artwork. • A video can only be a still. It is paused before the capture so the preview and the PDF agree; the settings page can blank it instead. • A page taller or wider than 60,000 CSS pixels is captured up to there and no further, because an infinite-scroll page has no bottom. The editor labels such a capture "Cut short" and says how much is missing before you export it.