Capture any page from top to bottom in a real browser. No watermark, no extension, no signup.
Free, no login, no email. Nothing stored beyond a 24 hour cache. Markdown copy · API and MCP
Paste any URL
We open the page in a real headless browser at a desktop viewport, the same way a visitor would see it.
We scroll the whole page
The browser scrolls to the bottom before capturing, so images and sections that only load when they come into view are actually in the shot.
Download the PNG
You get one image of the entire page. Download it straight away, because the link we give you stops working after about 24 hours.
Search for a full page screenshot tool and most of what you find takes payment in a currency other than money. A watermark in the corner, so the image is unusable in a client deck. A height cap that quietly cuts your landing page off at 8,000 pixels. An email gate before the download. A browser extension that wants permission to read every site you visit, forever, in exchange for a button you press twice a month. None of that is about the cost of taking a screenshot. It is about the cost of building an audience out of people who needed one.
This tool has none of those. You paste a URL and you get a PNG. The one thing it does ask of you is attention to a single fact: the link to that image expires in about a day, because the file lives in our storage bucket and we do not keep strangers' page captures forever. Download the file and it is yours. Leave the link in a document and it will be dead by next week. We say that next to the download button rather than in a footnote, because a screenshot that vanishes from a client proposal is a worse outcome than a watermark.
Capturing a whole page sounds like it should be one instruction to a browser, and it used to be. Then lazy loading became the default. Images below the fold now ship as empty placeholders that only fetch when they scroll into view, sections animate themselves in on an intersection observer, and half the page does not exist until somebody looks at it. Point a naive capture tool at a modern marketing site and you get the hero, then thousands of pixels of grey boxes and blank space where the content is supposed to be.
Getting it right means driving the browser the way a person would. Load the page, scroll all the way to the bottom, give the images that this triggers a moment to arrive, wait for the network to go quiet, then capture the full height in one pass. That is why this takes a few seconds rather than being instant, and it is the difference between a screenshot of a page and a screenshot of a page's loading state.
Design review, mostly. A full page capture is the only way to put the whole flow in front of somebody who is not going to click through it. Then there is the archive case: proving what a page said on a given day, before a price changed or a claim was edited. Competitive research runs on it too, because a folder of full page captures shows how a rival's messaging moved over a quarter in a way no notes ever do. And the quiet one is regression: capturing your own pages before and after a release, and putting the two side by side to see what moved that nobody meant to move.
About 24 hours. The capture is stored in our own cloud bucket and the link we hand you is signed with an expiry, so it stops working roughly a day after the capture. Download the PNG if you need to keep it, or paste the URL again later to take a fresh capture. We show the remaining time next to the download button so it is never a surprise.
The same run this page does, as an HTTP endpoint and as the MCP tool capture_full_page_screenshot. No account, no API key. Full reference.
curl -sS https://usesuperflow.ai/api/tools/full-page-screenshot \
-H 'Content-Type: application/json' \
-d '{"url":"example.com"}'Returns { imageUrl, expiresAt, bytes, width, height, deviceType }. 10 runs per hour per IP. Allow up to 90 seconds.
claude mcp add --transport http superflow https://usesuperflow.ai/api/mcpThen ask your agent for capture_full_page_screenshot. Setup for Claude Desktop, Cursor and VS Code is on the reference page.
No login, no email, no ads. Same as this one.
Superflow agents open your pages on a schedule, capture what they see, and tell you when a layout breaks, an image stops loading, or a section quietly disappears.