Convert a webpage to PDF
Save any web page as a PDF.
Enter a web address
Processed securely and deleted within the hour.
A real browser renders it
The page is loaded in Chromium and printed, which is the same engine and the same code path as your own browser's Save as PDF. That matters because a converter that parses HTML itself has to reimplement CSS, and modern layout — grid, flexbox, custom properties, web fonts, container queries — is not something you can approximate. Pages come out looking like pages.
JavaScript runs, so content that is rendered client-side appears. Backgrounds and background images are printed by default, which is the opposite of most browsers' print dialogue and almost always what someone converting a page actually wants.
What you can and cannot capture
Anything publicly reachable works. Anything behind a login does not — the conversion happens from a server with no access to your session, so a page that needs you signed in comes back as the login screen. There is no way around that other than printing to PDF from your own browser, where your session lives.
Pages that block automated visitors, sit behind a CAPTCHA, or serve different content to a datacentre address will produce whatever they show that visitor, which may be a challenge page.
Internal addresses are refused on purpose. Localhost, private ranges and cloud metadata endpoints are all blocked, because a URL fetcher that will retrieve them is a tool for reading things it should not — the class of problem known as server-side request forgery.
Why the result looks different from the screen
Because you are getting the page's print layout, and most sites have one. A print stylesheet commonly hides navigation, sidebars, cookie banners and adverts, expands collapsed sections, and switches to a serif face. That is normally an improvement — it is the version of the page designed to be read on paper.
It also means sticky headers may repeat, and elements positioned relative to the viewport may land oddly, because print has pages rather than a scrolling window.
Paper size and orientation
A4 by default, with Letter, Legal and A3 available. Landscape is worth reaching for on anything with wide tables or code blocks, which otherwise get cut off at the page edge — a portrait page is about 80 characters wide at a normal font size, and a line of code is frequently longer.
What survives into the file
Text stays text, so the result is searchable, selectable and copyable rather than a picture of a page. Links stay clickable. Vector content stays vector and prints sharply at any size.
What does not survive: video, animation, anything requiring interaction, and anything that only appears after scrolling if the site loads it lazily. A page built as an infinite feed will give you the first screenful or two.
Common questions
Can it capture a page I have to log in to?
No. The conversion runs from a server with no access to your session, so you get the login screen. Print to PDF from your own browser for those.
Why does the PDF look different from my screen?
You are getting the site\u2019s print layout, which usually hides navigation, sidebars and adverts and expands collapsed sections. That is generally an improvement.
Is the text selectable?
Yes. A real browser renders it and text stays text, so the result is searchable and copyable rather than an image of the page.
Do backgrounds and images print?
Yes, by default — the opposite of most browsers\u2019 print dialogue, and almost always what you want when converting a page deliberately.
Why was my URL rejected?
Internal addresses are blocked on purpose: localhost, private network ranges and cloud metadata endpoints. A fetcher that will retrieve those is a tool for reading things it should not.
My wide table is cut off.
Switch to landscape. A portrait page fits roughly 80 characters at a normal size, which is narrower than most tables and many lines of code.
Does JavaScript run?
Yes, so client-rendered content appears. Content that only loads as you scroll may not, since printing does not scroll.