If you have ever taken a "full page" screenshot of Gmail, a Notion doc or an analytics dashboard and got back one screen of content with a lot of empty space, this post explains why, and what to do about it.
Disclosure up front: I build FullPage Capture, a Chrome extension for this. The first half of the post needs no extension at all.
The built-in way, and where it stops
Chrome can capture a whole page without installing anything:
- Open DevTools (F12, or Cmd+Option+I on Mac)
- Press Ctrl+Shift+P (Cmd+Shift+P on Mac)
- Type
screenshotand pick Capture full size screenshot
For articles, docs sites and landing pages this is all you need. It fails in four specific situations.
1. The page does not scroll, a panel inside it does
Gmail, Notion, Slack, most admin dashboards and many editors keep the window fixed and scroll a div inside it. DevTools measures the document height, finds it equals the viewport, and saves one screen. The content you wanted is inside a container the capture never scrolled.
You can check this yourself: if the browser scrollbar on the far right is missing but the content still scrolls, it is an inner scroller.
2. Sticky headers
Tools that scroll and stitch take a picture per screen. A header with position: sticky or fixed is present in every one of those pictures (floating chat buttons and cookie bars behave the same way), so it shows up again every 800 pixels or so in the result. The fix is to capture the fixed elements once and hide them for the remaining frames.
3. Lazy-loaded content
Images and list rows that load on scroll are not there until something scrolls past them. A capture that resizes the viewport instead of scrolling gets grey placeholders. You have to scroll, wait for the network, then shoot.
4. Very long pages
Chrome limits how large a single canvas can be. Past that limit the image silently comes out blank or truncated. Long chat exports and infinite feeds hit this regularly.
What I do about each one
FullPage Capture handles these cases like this:
- Inner scrollers and iframes: it captures the window, then finds the scrolling panel and unrolls it too.
- Sticky headers: captured once, at the top.
- Lazy loading: it scrolls the real page and waits before each frame.
- Very long pages: past 28,800 px per side the capture is split into numbered parts instead of failing.
What people tell me they use it for:
- Sending a complete page as documentation to an agency, a client or a government portal. Capture the whole page, upload one file.
- Grabbing website layouts and content for design reference.
- Saving a ChatGPT or Claude conversation, a Gmail thread or a Notion page as a PDF.
- Capturing a full Google Analytics or Search Console report.
- Attaching a complete page to a bug report with the URL and date stamped on it.
A few lines from recent Chrome Web Store reviews, quoted as written:
"It's the best way to provide online documentation to agencies. Just click the whole page and upload it to their site."
"My go to for grabbing website layouts and content."
"A rare one that does long page screen capture easily."
Not every review is glowing. One user reports that very long pages can come out blurry, and another that captures lag on heavily interactive sites. Both are on my list.
Export is PNG, JPEG or PDF. The PDF can be one tall page or Letter, Legal and A4 with page breaks that do not cut through a line of text, and links stay clickable.
Where other tools are the better choice
- You use Firefox. FullPage Capture is Chromium only. Firefox has a good built-in full page screenshot, and GoFullPage now ships a Firefox version.
- You want zero install permissions. GoFullPage requests no site access at install. FullPage Capture asks for it, because reading iframes and PDFs open in Chrome requires it.
- You capture ordinary pages only. On plain document-scrolling pages the DevTools command and every major extension give the same result. Use whichever is already installed.
Privacy and price
Captures are processed and stored in the browser. There is no account and no analytics. One exception worth stating: extensions cannot scroll Chrome Web Store pages, so for those the public page address is sent to our server, which renders it and returns the image.
Capturing, saving, copying and printing are free with no watermark. Saving a capture after you have annotated it, searchable PDF, batch capture and an evidence stamp with a SHA-256 checksum are in Pro at $19 a year.
- Site and guides: https://fullpagecapture.net
- Chrome Web Store: https://chromewebstore.google.com/detail/fullpage-capture-full-pag/ggacghlcchiiejclfdajbpkbjfgjhfol
If you have a page that still captures wrong, post the URL in the comments. Broken cases are how the capture logic gets better.
Top comments (1)
The canvas ceiling you mentioned (
28,800 pxsplit threshold) is a common surprise when people build DOM-to-PDF exporters. Depending on the GPU backing store, Chromium hits max canvas texture or total pixel-area limits well before long Notion docs finish unrolling, andtoDataURL()simply returns transparent pixels without throwing.One detail that helps when exporting multi-page PDFs from stitched frames: pure raster slices drop selectable text and break hyperlinks. Extracting link bounding boxes with
getClientRects()relative to the scroll container, multiplying by page scale, and writing/Linkannotation rectangles onto each PDF page keeps URLs clickable even across split slices.