Website Screenshot & PDF: Full Page, Most Cookie Banners Hidden
Pricing
$2.00 / 1,000 screenshot or pdf files
Screenshots (PNG, JPEG, WebP) and PDFs of web pages or HTML. Full page with lazy images, mobile presets, common cookie banners hidden. You pay only for files produced.
The web pages to capture, one file per address. Only http and https addresses are captured. An address that is not valid, or is not http/https, gives a free record with the reason; a page that answers an error (404, 500...), refuses (401, 403, 429), asks for a challenge, does not load in time, or points to a private network address gives a free record too. Use either this list, or the HTML field below, or both. Lists of addresses in a remote file ("requestsFromUrl") are not supported.
A complete HTML page to render instead of (or besides) the addresses, for example an invoice or a report to turn into a PDF. Styles must be inline or in a tag, and images must be data: addresses or absolute https addresses: relative addresses cannot be resolved. Up to 5 MB. The record has inputType "html" and no url.
png (lossless, default), jpeg and webp (smaller, use Quality) or pdf. One file per address, stored in the key-value store of the run.
Capture the whole page, not only the first screen. Ignored for PDF (a PDF always contains the whole page) and when a selector is given. Pages taller than the maximum height are cut there and marked truncated: true.
Tallest capture in CSS pixels, from 100 to 16384 (the limit of Chromium). A page taller than this is cut at this height and the record says truncated: true and gives fullHeightPx. WebP cannot be taller than 16383 pixels, so 16383 is its cap.
From 1 to 100. No effect on PNG and PDF.
The screen to emulate. Desktop is 1440 x 900. The phone and tablet presets also send a mobile browser identity and touch. Width and Height below override the size of the preset.
Overrides the width of the device preset, from 200 to 3840. The record then says device: "custom".
Overrides the height of the device preset, from 100 to 4320. It is the height of the first screen; with Full page the capture goes beyond it.
1 to 3. 2 gives a sharp retina image twice as wide and tall (4 times the pixels, so more memory and bigger files). Default 1.
Light or dark, for sites that follow the system setting (prefers-color-scheme).
For example #pricing-table or .hero. Only that element is captured (the first one that matches). If nothing matches, the record is empty_capture and free. Ignored for PDF.
When the page counts as loaded. load (default) waits for the load event; domcontentloaded is faster; networkidle also waits for the network to be quiet, at most 15 seconds more (a page that never gets quiet is still captured).
Wait until this element is in the page (up to 20 seconds) before capturing. If it does not show up the record is timeout and free.
Wait this long after the page has loaded and scrolled, for animations or late content. 0 to 10000.
Before a full page, PDF or element capture, scroll down the page step by step so that lazy-loaded images and sections load, then go back to the top. Capped at 12 seconds and at the maximum height, so an endless page cannot hang the run.
Elements to hide before the capture, for example chat widgets or sticky bars. Up to 50 selectors.
CSS added to the page before the capture, up to 20,000 characters.
Do not load requests to a short, fixed list of advertising networks (about 25 domains, such as doubleclick.net and googlesyndication.com). It is not a general tracker blocker.
Paper size when the format is pdf.
For example {"top": "10mm", "right": "10mm", "bottom": "10mm", "left": "10mm"} (the default). Units: mm, cm, in or px.
{ "top": "10mm", "right": "10mm", "bottom": "10mm", "left": "10mm"}Print the backgrounds of the page (on by default; browsers drop them from printouts otherwise).
print (default) uses the print styles of the page, as a browser Save as PDF does, and respects @page rules (good for invoices and articles). screen keeps the look of the screen.
For example {"Authorization": "Bearer ..."}. Sent only to the address of each page itself, never to the third-party resources it loads. Host, Cookie, User-Agent and similar headers are ignored. Never written to the log or to the output.
For example {"username": "user", "password": "..."}. Sent only to the origin of each page. Never written to the log or to the output.
A page that has not been captured in this time is given up (record timeout, free) so that the run never hangs. A timeout, a reset connection or a 5xx answer is tried once more. From 2 to 120.