TL;DR
To download a news article as a PDF: open it in Chrome, click Convert: Web to PDF, toggle Article Mode to strip the navigation and ad rails, and download. Ten seconds, no account, no watermark, and the output is a real PDF with selectable text and clickable links rather than a screenshot.
But the reason to do it isn't offline reading. It's that news articles are living documents that change after publication and almost never say so. Headlines get rewritten for A/B tests. Numbers get corrected. Quotes get softened after a subject complains. Paragraphs get added as a story develops. Sometimes the whole piece gets pulled.
If you cite a news article in a paper, a legal filing, a compliance record, or an internal analysis, you're citing a version — and the URL doesn't preserve versions. The PDF you save is the only artifact that says what the article said on the day you read it.
This post is about that: the versioning problem, what a good news-article PDF needs to contain to be worth anything later, and the specific mechanics of getting it out of modern news sites cleanly.
The problem: URLs point at the current version, not the one you read
Every citation format assumes a stable target. You give a URL and an accessed-date, and the reader is supposed to be able to check your work.
That assumption held reasonably well for print and held badly from the moment news moved online. Consider the ordinary lifecycle of a breaking story:
- Minute 0. Two hundred words go up with a provisional headline. No byline yet, or a wire credit.
- Hour 1. The piece triples in length. The headline changes twice as the desk tests framing.
- Hour 6. A named reporter's byline replaces the wire credit. The lede is rewritten around a new quote.
- Day 2. A correction is folded into the body. Depending on the outlet, there may or may not be a correction note.
- Week 2. The story is updated with the outcome, effectively becoming a different article at the same URL.
At no point in that sequence does the URL change. If you read it at hour one and cite it at week two, your citation points at something you never saw.
Which changes get disclosed, and which don't
Standards vary enormously, and it's worth knowing where the lines usually fall:
| Change type | Typically disclosed? | Visible in the page later? |
|---|---|---|
| Factual correction | Often, at reputable outlets | Sometimes, as a correction note |
| Headline rewrite | Almost never | No |
| Quote trimmed or removed | Rarely | No |
| Paragraph added for developing story | Sometimes ("Updated at…") | Only as a timestamp |
| Photo or caption swap | Almost never | No |
| Article unpublished entirely | Varies | No — you get a 404 |
The two rows that cause the most trouble in practice are headline rewrites and quote changes, because they're the ones people quote most and the ones least likely to leave a trace.
Who actually needs this
Students and academics. A citation to a news article is a citation to a claim. If your argument rests on "the Financial Times reported X on July 12," and by the time your reader checks the article it reports X′, the burden of explanation is on you. A PDF in your reference folder, named with the date, resolves it in one move.
Legal and compliance teams. Media monitoring is a standard part of regulatory and reputational work. "We reviewed coverage as of this date" is only meaningful with the coverage attached. Our post on capturing webpages as evidence honestly is candid about the limits — a self-made PDF is not a notarized capture — but it's dramatically better than a dead link.
Analysts and researchers. If you're tracking how a narrative shifted, you need the versions, not the endpoint. The endpoint is exactly the thing that hides the shift.
Anyone doing FOIA, journalism, or accountability work. Stories get pulled. Sometimes for good reasons, sometimes not. Either way, an unpublished article you didn't archive is an argument you can no longer make.
What a useful news-article PDF has to contain
Most people capture too little. A clean-looking PDF of just the body text is pretty and half-useless. Make sure your capture includes:
- The full headline and any subheadline or standfirst. The headline is the most-changed and most-quoted element.
- The byline. Wire credit versus named reporter is meaningful, and it changes.
- The publication timestamp and any "updated" timestamp. Both. The gap between them is information.
- The full body, including anything behind a "read more" expander.
- The URL. Some sites print it in the footer of a print stylesheet; if not, note it in the filename or annotate the PDF afterward.
- Any correction or editor's note, usually at the very bottom.
- Embedded quote cards, pull quotes, and social embeds where they carry content rather than decoration.
Article Mode's Readability pass is very good at 1, 2, 4 and 7 and occasionally clips 3 and 6, because timestamps and correction notes sometimes sit outside the detected article body. Check the bottom of your PDF for the correction note before you close the tab. If it's missing, redo the capture in default mode.
The mechanics, site by site pattern
Modern news sites share a handful of layout patterns, and each has a specific failure mode.
Standard article page
Article Mode handles this cleanly. Click the extension, toggle Article Mode, preview, download. You'll get headline, byline, body, and images without the nav, the newsletter interstitial, the "most read" rail, or the sticky subscribe bar.
Live blog / developing story
Article Mode is the wrong tool here. Live blogs are a reverse-chronological stack of timestamped entries; Readability tends to grab one entry and discard the rest. Use default mode, and scroll the full page first so lazy-loaded entries render. Live blogs can run to hundreds of entries — expect a long PDF, and that's correct.
Image-heavy feature or photo essay
Use Load All Images before converting. Feature pages defer image loading aggressively, and without pre-loading you'll get a PDF full of gaps where the photography should be. This is the single most common complaint about webpage-to-PDF output and it has a one-toggle fix.
Article with an embedded data visualization
Charts rendered as SVG or canvas generally capture fine. Charts rendered in an iframe from a third-party host may not, because the iframe content is a separate document. If a chart is essential, capture it separately with Capture Element, or screenshot it and note the source.
Paywalled article
Be direct about this: the extension does not bypass paywalls. It converts what your browser is already displaying. If you're a subscriber and logged in, the full article is on your screen and it converts normally — that's the whole point of running locally rather than on a server, since a server-side converter fetching the public URL only ever sees the paywall.
If you're not a subscriber, you'll get whatever the free preview shows, and no tool we make will change that. The extension's FAQ covers the specific behavior on Substack and Medium member overlays, which are overlays over already-delivered content rather than true paywalls — a different situation with a different answer.
Code-heavy technical article
If the piece contains code blocks, skip Article Mode and use default mode. The Readability algorithm sometimes drops <pre> blocks when isolating the article body, and a tutorial without its code is worse than useless. Default mode plus a few Remove Elements clicks on the sidebar gets you a clean result with the code intact.
Naming files so future-you can find them
This is the part everyone skips and everyone regrets. A folder of article.pdf, article (1).pdf, article (2).pdf is not an archive.
A convention that works:
2026-07-28_reuters_chip-export-controls_v1.pdf
- Date first so the folder sorts chronologically.
- Outlet second so you can filter by source.
- Slug third so you can search by topic.
- Version suffix if you're deliberately tracking a developing story.
If you're capturing the same story repeatedly as it develops, v1, v2, v3 on the same date-outlet-slug stem gives you a diffable set. Open two side by side and the changes are obvious in seconds.
Comparing the options honestly
| Method | Selectable text | Links work | Handles lazy images | Works when logged in | Uploads your page |
|---|---|---|---|---|---|
| Convert: Web to PDF | Yes | Yes | Yes, with Load All Images | Yes | No |
| Chrome Print to PDF | Yes | Yes | Often not | Yes | No |
| PrintFriendly | Yes | Yes | Partly | Limited | Some operations server-side |
| GoFullPage → PDF | No — flat image | No | Yes | Yes | Varies |
| PDFCrowd / URL converters | Yes | Yes | Varies | No | Yes |
| Adobe Acrobat online | Yes | Yes | Varies | No | Yes |
| Reader Mode → Print | Yes | Yes | Varies | Yes | No |
| Browser bookmark | n/a | n/a | n/a | n/a | Preserves nothing |
Two columns decide this. "Works when logged in" rules out every URL-based converter for subscriber content — they fetch the page from their own servers and get the logged-out version. "Uploads your page" matters if you're archiving anything sensitive; a local extension makes no network request, so there's nothing to leak.
The one genuinely competitive alternative is Chrome's own Print to PDF, which is free and already installed. It's fine for simple pages. It falls over on lazy-loaded images, sticky headers that repeat on every page break, and cookie banners baked into the output. We compared them properly in Chrome Print to PDF vs. a dedicated extension.
Building an actual archive habit
Three rules that make the difference between a folder of PDFs and a usable archive:
- Capture at read time, not later. The article you meant to save is the one that got rewritten. Ten seconds now beats reconstruction never.
- Capture the version, not the topic. If a story is developing and you care about it, capture again when it materially changes. Two snapshots a week apart tell a story one snapshot can't.
- Write the URL somewhere machine-readable. Filename, PDF annotation, or a companion text file. A PDF with no source URL is an orphan.
If you're monitoring dozens of outlets rather than saving the occasional article, the manual approach doesn't scale and you want the structured version instead: ScrapeMaster can read a search or topic page and pull headline, byline, timestamp and URL into a CSV, giving you a change log across many stories. Then PDF only the ones where the log shows something moved. That's covered in scraping for research territory and pairs well with this workflow.
Frequently asked questions
How do I download a news article as a PDF?
Open the article in Chrome, click Convert: Web to PDF, toggle Article Mode to remove navigation and ads, then download. The result is a real PDF with selectable text and working links. It takes about ten seconds and requires no account or upload.
Why not just use Chrome's Print to PDF?
It works for simple pages. It struggles with the things news sites do constantly: lazy-loaded images that never render, sticky headers that repeat on every page break, cookie banners captured into the output, and interstitials that appear mid-page. Article Mode plus Load All Images solves those specifically.
Do news articles really change after publication?
Routinely. Headlines are rewritten, sometimes several times in the first hours. Corrections get folded into the body. Quotes get trimmed. Developing stories are appended to for days. Most outlets disclose factual corrections; almost none disclose headline changes or quote edits, and the URL never changes.
Can this download articles behind a paywall?
No. The extension converts what your browser is already showing. If you're a logged-in subscriber, the full article is on screen and converts normally — which is exactly why a local extension beats a server-side converter, since the server only ever sees the logged-out page. If you're not a subscriber, you'll capture the free preview and nothing more.
Will the PDF include the correction note at the bottom?
Usually, but check. Article Mode isolates the detected article body, and correction notes sometimes sit outside it. If the note is missing from your PDF, redo the capture in default mode and use Remove Elements to clear the sidebar instead.
How do I capture a live blog or developing story?
Use default mode rather than Article Mode — Readability tends to grab a single entry from a reverse-chronological stack. Scroll the whole page first so lazy-loaded entries render, then convert. Expect a long PDF; that's the accurate output.
What's the best way to name saved article PDFs?
Date, outlet, slug, version: 2026-07-28_reuters_chip-export-controls_v1.pdf. Date-first sorts chronologically, and the version suffix lets you keep multiple captures of a developing story in one diffable set.
Is anything uploaded when I convert an article?
No. Conversion runs locally through Chrome's own print engine. The page content, URL, cookies, and session never leave your machine. That's what makes the extension usable on subscriber content and internal tools alike.
Bottom line
A news URL is a pointer to the current version. If your work depends on what an article said rather than what it says, you need the version, and the only way to get it is to capture it when you read it.
Article Mode, Load All Images, a sensible filename, and a check that the correction note made it in. Ten seconds per article, and it turns a link that might 404 into a document you can actually cite.
Convert: Web to PDF is free, runs entirely in your browser, adds no watermark, and never uploads what you're reading.
And when the reading's done: CineMan AI puts IMDb and Rotten Tomatoes ratings directly on Netflix, Prime Video, and Disney+ so picking something takes thirty seconds instead of twenty minutes.