Webpage to PDF
Save an article as a clean, readable PDF — without the navigation, adverts and sidebars.
Extracted
Two different jobs
"Save a web page as a PDF" turns out to mean two quite different things, and it is worth knowing which one you want before picking a tool.
| This tool | Browser print to PDF | |
|---|---|---|
| Result | Clean reading copy, typeset | A picture of the page as designed |
| Navigation, ads, sidebars | Removed | Usually included |
| Images | Left out | Included |
| Text | Real, selectable, searchable | Real, selectable, searchable |
| File size | Tens of kilobytes | Often several megabytes |
| Pages behind a login | Cannot read them | Works, you are already signed in |
| Editing before saving | Yes | No |
For archiving a receipt or capturing a design, print to PDF. For keeping an article to read later, filing documentation, or turning a reference page into something you can annotate, the reading copy is smaller, tidier and far easier to search.
Why you can edit the text first
Extraction is a judgement, and no tool gets it right every time: a related-articles strip can look exactly like an article's own list of links. So what was pulled out is handed to you as editable Markdown before anything is typeset. Delete the stray menu item, cut the newsletter box, add a note at the top about why you saved it. Then convert.
How the page is read
Every other PDF tool on this site runs entirely in your browser. This one cannot, and the reason is worth stating rather than glossing over: a browser is not permitted to read a page from another site. The same-origin policy forbids it, and that rule is what stops any web page you visit from quietly reading your webmail.
So this tool has one server-side step and only one. Our server fetches the markup at the address you gave and returns the readable structure it found. Everything after that — the layout, the typesetting, the writing of the PDF — happens in your browser, as it does everywhere else here. The document you download is built on your machine and never existed on ours.
The fetch itself is guarded. A tool that will retrieve any address a stranger types is an invitation to use the server as a way into the private side of a network, so the address is checked against the private and reserved ranges, the connection is pinned to the address that was checked, and every redirect is put through the same tests rather than being followed blindly.
What works and what does not
- Works well: news articles, blog posts, documentation, reference pages, wikis, changelogs — anything that sends real HTML.
- Comes back thin: pages that draw themselves with JavaScript after loading. There is nothing in the delivered markup to read, and no tool outside a browser can read them.
- Will not work: anything behind a login, a paywall or a bot check. The page is fetched anonymously, so what comes back is what a stranger would see.
Questions
Does the PDF look exactly like the web page?
No, and that is deliberate. This produces a clean reading copy: the article text, headings, lists, quotes and tables, typeset on a page, without the navigation, adverts and sidebars. If you want a picture of the site as designed, use your browser's own Print command and choose Save as PDF, which uses the engine that drew the page in the first place.
Why does this tool need a server when the others do not?
Because a browser is not allowed to read a page on another site. The same-origin policy blocks it, so something has to fetch the markup on your behalf. Our server does only that one step and passes the readable text back; the PDF itself is still built in your browser.
Can it read pages behind a login?
No. The page is fetched anonymously, so anything requiring a sign-in, a paywall or a cookie comes back as the version a stranger would see. For those, print to PDF from the browser where you are already signed in.
Some pages come back nearly empty. Why?
Those pages build themselves with JavaScript after loading, so the markup delivered to a fetch is close to blank. Nothing running outside a browser can read them. Sites that send real HTML — most news, documentation, blogs and reference pages — work well.
Are the images included?
No. The result is a typeset text document, so pictures are left out and figure captions are kept in their place. That keeps the file small and every word of it selectable and searchable.
Can I fix the text before converting?
Yes. What was extracted is shown as editable Markdown before anything is typeset, so a stray menu item or a duplicated heading can be deleted, and you can add your own notes to the top.
Related tools
Browse all Sigma Wire tools - every free tool on the site, grouped by category.