TL;DR
Search the Chrome Web Store for a webpage-to-PDF plugin and you get dozens of listings with near-identical copy: one click, high quality, free, no watermark. The listings don't distinguish them because the listings aren't where the differences live.
Seven tests do distinguish them, and each takes about a minute on a page you already have open. Run them before you commit to a tool:
- Is the output real text, or a screenshot? Try to select a word in the PDF.
- Do lazy-loaded images make it in? Test on an image-heavy article.
- Does it work on a page you're logged into? Test on your own email or a dashboard.
- Does anything leave your machine? DevTools → Network during a conversion.
- Can you remove elements before exporting? Try to delete a cookie banner.
- Does it handle a long page in one pass? Test on something 30 screens deep.
- What does it request permission to do? Read the disclosure, check site access.
We make Convert: Web to PDF, so run these on ours too. We'd rather you test than take our word for it — and where a test exposes a limit of ours, this post says so.
Test 1: Real PDF or screenshot in a PDF wrapper?
The test: convert a page, open the PDF, and try to select a sentence with your cursor. Then Cmd+F and search for a word you can see.
Pass: the text highlights, the search finds it, and you can copy-paste out of it.
Fail: nothing selects. You have an image inside a PDF container.
Why it matters more than it sounds. Screenshot-based PDFs are not searchable, not quotable, not machine-readable, and can't be indexed by anything. Links inside them are dead pixels. And they can't be made accessible — screen readers have nothing to read, and no amount of tagging fixes an image. If accessibility conformance matters to you, a screenshot PDF is a dead end that requires full re-authoring, while a real text PDF is a starting point.
Several popular full-page-capture extensions are screenshot tools with PDF export bolted on. GoFullPage is the well-known example — excellent at what it is, which is a screenshot tool. Converting its output to PDF gives you a picture of a page, not a document. We compared them directly in GoFullPage vs. Convert: Web to PDF.
Test 2: Lazy-loaded images
The test: find a long, image-heavy page — a photo essay, a long product listing, a documentation page with lots of diagrams. Don't scroll. Convert immediately. Then scroll to the bottom, and convert again.
Pass: both PDFs contain the images.
Fail: the first is full of blank gaps where images should be.
Why it matters. Almost every modern site defers image loading until an image approaches the viewport. A tool that captures the current DOM state without pre-loading gets the placeholders. This is the single most common complaint about webpage-to-PDF output, and it has an unglamorous fix: the tool needs to pre-scroll the page, or offer a "load all images" step, before capturing.
Chrome's built-in Print to PDF fails this test on most sites, which is the main reason people go looking for an extension in the first place. We wrote up the specific fix in fixing missing images in webpage PDFs.
Test 3: Logged-in pages
The test: open something that requires authentication — your webmail, a bank statement view, an internal dashboard, an order history. Convert it.
Pass: the PDF contains what you were looking at.
Fail: the PDF is a login screen, or an error, or empty.
Why it matters. This test cleanly separates tools by architecture. An extension that runs locally reads the page your authenticated session already rendered — it sees exactly what you see. A tool that sends the URL to a server for rendering has no session on that site, so it fetches the logged-out version.
That's why URL-based converters — PDFCrowd, Adobe's online webpage-to-PDF, and every "paste a link" service — cannot do this at all. It's not a quality difference; the approach structurally can't work. Some Chrome extensions are thin wrappers around exactly such a service, and this test exposes them immediately.
If everything you convert is public, this may not matter to you. If any of it is an invoice, a dashboard, a portal, or a receipt, it's the whole game.
Test 4: Does anything leave your machine?
The test: open DevTools (F12 or Cmd+Option+I), go to the Network tab, clear the log, then run a conversion and watch.
Pass: you see the page's own requests and nothing that looks like your content being transmitted.
Fail: a POST to an unfamiliar host carrying page content or a document body.
Why it matters. "Private" and "local" are unregulated words in a store listing. This test is the only way to check. Ninety seconds, and it works on any extension.
Be fair when reading the results. A single small analytics ping isn't the same as uploading your document — plenty of tools count usage. What you're looking for is your content going somewhere. Our own extension sends one anonymous ping after a successful conversion containing a random install token and nothing else — no URL, no content — which is disclosed in the privacy policy and visible in this test. We mention it because you'll see it, and an unexplained request is worse than a disclosed one.
The broader method is in auditing Chrome extension privacy disclosures.
Test 5: Removing elements before export
The test: find a page with a cookie banner, a newsletter interstitial, or a sticky header. Try to remove it before converting.
Pass: you can click elements away, ideally with an undo, or switch to a mode that strips everything but the main content.
Fail: whatever's on the page is in the PDF, cookie banner and all.
Why it matters. This is the difference between a PDF you'd send to someone and one you'd be slightly embarrassed by. Sticky headers are the worst offender — they repeat on every single page break, eating a strip off every page.
Two mechanisms worth having, because they solve different problems:
- Click-to-remove individual elements, with undo. Precise, works anywhere, needed when a page has one specific annoyance.
- Article Mode (a Readability-based reader view) that strips everything except the main content. Fast, ideal for articles, and blunt — it decides what "main content" means, and on a structured page like a dashboard or a settings screen it will pick one region and discard the rest.
Both is better than either. And know Article Mode's limit: it can drop <pre> blocks, so for code-heavy pages use the default mode with manual removal instead.
Test 6: Long pages in one pass
The test: find something genuinely long — a lengthy documentation page, a long thread, a 5,000-word article. Convert it.
Pass: one PDF, all the content, sensible page breaks.
Fail: it truncates, times out, or produces one enormous unusable page.
Why it matters. Plenty of tools are fine at one screen and fall over at thirty. Watch specifically for content cut off partway, and for page breaks landing mid-sentence or mid-image on every page.
Infinite scroll is a genuine limitation, not a bug, and any honest tool will tell you so. A page that loads more content as you scroll has only fetched what you've scrolled to. No extension can capture content the page hasn't requested. The workflow is: scroll to the endpoint you want, then convert. Anything claiming to capture an entire infinite feed in one click is overselling.
Test 7: Permissions
The test: before installing, read the Chrome Web Store data disclosure. After installing, go to chrome://extensions → Details → Site access.
Pass: permissions are proportionate, and site access can be set to "On click."
Fail: broad host permissions with no explanation, or a disclosure that doesn't match the behaviour.
Why it matters. A webpage-to-PDF extension needs to read the current page. It does not need persistent access to every site you visit. "On click" activation is dramatically narrower than "On all sites," and most conversion tools work fine with it.
Two things to note honestly. First, some of these tools request the debugger permission, which sounds alarming and is how Chrome exposes its DevTools Protocol — including Page.printToPDF, the API that produces real text PDFs rather than screenshots. It's the legitimate mechanism for high-quality output. It's also a powerful permission, which is exactly why Test 4 matters: verify the behaviour rather than trusting the label in either direction.
Second, ownership changes are the main way a good extension goes bad. An extension that's been fine for three years can be sold and updated with new behaviour. Check who publishes it, and treat a change of developer as a reason to re-run these tests.
The results table
| Capability | Convert: Web to PDF | Chrome Print to PDF | PrintFriendly | GoFullPage | PDFCrowd / URL-based | Adobe online |
|---|---|---|---|---|---|---|
| Real selectable text | Yes | Yes | Yes | No — image | Yes | Yes |
| Lazy-loaded images | Yes | Often not | Partly | Yes | Varies | Varies |
| Logged-in pages | Yes | Yes | Limited | Yes | No | No |
| Fully local | Yes | Yes | Partly server-side | Varies | No | No |
| Remove elements | Yes, + undo | No | Yes | No | No | No |
| Article/reader mode | Yes | No | Yes | No | Some | No |
| Capture one element | Yes | No | No | Region select | No | No |
| Paper sizes | A3–Ledger | Several | Limited | n/a | Several | Several |
| Watermark-free | Yes | Yes | Yes | Yes | Free tiers often not | Account required |
| Cost | Free | Free | Freemium | Freemium | Freemium | Subscription |
Chrome's built-in deserves genuine credit: it's free, already installed, produces real text PDFs, and handles simple pages perfectly well. If your pages are simple, install nothing. The extension category exists for the pages that fight back — lazy images, sticky headers, cookie banners, and content you need cleaned up before sending.
What none of them do
Worth stating so you can stop looking:
- No OCR. If a page shows a scanned document as an image, the PDF contains that image and its text stays unsearchable. Nothing in this category runs text recognition.
- No paywall bypass. These convert what your browser is displaying. If you can't see it, neither can they.
- No editing existing PDFs. These make PDFs. They don't reflow, redact, or re-paginate one you already have.
- No Firefox or Safari. The relevant APIs are Chromium-specific. Chrome, Edge, Brave, Arc, Opera, and Vivaldi work; the others don't.
Frequently asked questions
What should I look for in a Chrome extension that converts webpages to PDF?
Real selectable text rather than a screenshot; correct handling of lazy-loaded images; the ability to work on pages you're logged into; fully local conversion with nothing uploaded; element removal before export; reliable handling of long pages; and permissions proportionate to the job. Each is testable in about a minute.
How can I tell if an extension makes a real PDF or just a screenshot?
Open the output and try to select text with your cursor, then use Cmd+F to search for a visible word. If nothing selects and search finds nothing, it's an image inside a PDF container — not searchable, not quotable, and not fixable for accessibility without re-authoring.
Why do images go missing from my webpage PDFs?
Modern sites defer loading images until they approach the viewport. A tool that captures the page as-is gets placeholders instead. The fix is a tool that pre-scrolls the page or offers a "load all images" step before capturing — or scrolling to the bottom yourself first.
Can a Chrome extension convert a page I'm logged into?
A local one can, because it reads the page your authenticated session already rendered. A URL-based service cannot, because it fetches the page from its own servers with no session on that site and receives the login screen instead. This is the clearest architectural difference between the two approaches.
Is the "debugger" permission dangerous?
It's powerful, and it's the legitimate route to Chrome's DevTools Protocol — including the printToPDF API that produces real text PDFs instead of screenshots. Rather than judging the permission label, verify the behaviour: open DevTools' Network tab, clear it, run a conversion, and see whether anything leaves your machine.
How do I check whether an extension uploads my pages?
Open DevTools with F12, go to the Network tab, clear the log, and run a conversion. Watch for any request carrying page content to an unfamiliar host. Distinguish a small anonymous usage ping from your content being transmitted — the first is common and often disclosed, the second is the thing to avoid.
Can any extension capture an entire infinite-scroll page in one click?
No, and any claiming to is overselling. A page that loads content as you scroll has only fetched what you've scrolled to; there's nothing in the DOM to capture beyond that. Scroll to the endpoint you want first, then convert.
Is Chrome's built-in Print to PDF good enough?
For simple pages, yes — it's free, installed, and produces real text PDFs. It struggles with lazy-loaded images, sticky headers repeating on every page break, cookie banners baked into the output, and it offers no way to remove elements first. That gap is what the extension category exists to fill.
Bottom line
Store listings all say the same things because the differences aren't visible in a listing. Seven tests — text selection, lazy images, logged-in pages, network traffic, element removal, long pages, permissions — sort the category in under ten minutes, on pages you already visit.
Run them on whatever you're considering. Run them on ours. The tool that passes on your actual pages is the right one, and it's a much better basis for choosing than a star rating.
Convert: Web to PDF is free, has no account, adds no watermark, and runs entirely on your machine. If you'd rather read the specifics than test, the 117-question FAQ covers most of what comes up.
And for a decision that doesn't need seven tests: CineMan AI puts IMDb and Rotten Tomatoes ratings directly on Netflix, Prime Video, and Disney+.