PagedPDF vs Browserless for PDFs
Checked: prices and features from browserless.io and docs.browserless.io on 2026-11-11. We haven't run Browserless ourselves, so its side comes from its documentation.
These are different kinds of product. Browserless hosts headless browsers that you drive with Puppeteer or Playwright, for scraping, testing, screenshots and automation, and it has a /pdf endpoint among them. PagedPDF only turns HTML into PDFs.
At most volumes, Browserless costs less per PDF. What you're choosing is how much of the PDF work you want to keep doing yourself.
This page is written by PagedPDF. Check anything that matters to you against Browserless's pricing page.
How Browserless bills
Browserless sells units, not documents. In its words, a unit is "a block of browser time of up to 30 seconds per browser connection", and a longer session uses another unit for every 30 seconds. A PDF that renders in under 30 seconds is one unit, so the tables below count one unit per PDF.
| Browserless plan | Monthly billing | Yearly billing, per month | Units a month | Extra units |
|---|---|---|---|---|
| Free | $0 | $0 | 1,000 | None: requests are rejected |
| Prototyping | $35 | $25 | 20,000 | $0.0020 |
| Starter | $200 | $140 | 180,000 | $0.0017 |
| Scale | $500 | $350 | 500,000 | $0.0015 |
Cost per document
Monthly billing on both, taking the cheapest route, overage included:
| PDFs a month | Browserless | Each | PagedPDF | Each |
|---|---|---|---|---|
| 500 | $0 (Free) | $0 | $15 (Starter) | $0.03 |
| 2,000 | $35 (Prototyping) | $0.0175 | $15 (Starter) | $0.0075 |
| 10,000 | $35 (Prototyping) | $0.0035 | $49 (Growth) | $0.0049 |
| 50,000 | $95 (Prototyping plus 30,000 extra units) | $0.0019 | $149 (Scale) | $0.0030 |
PagedPDF is cheaper only between 1,000 and 2,000 PDFs a month. Below 1,000, Browserless is free. Above 2,000 it's cheaper, and the gap widens with volume.
The units are shared, though. If the same account also runs scrapers or tests, those use the same pool, and a slow PDF that takes 31 seconds costs two units.
What Browserless leaves to you
The /pdf endpoint takes Puppeteer's PDF options and prints with Chrome, so you get Chrome's defaults and Chrome's quirks. These are the ones we've tested in Puppeteer and written guides for:
- Table rows split across pages until your CSS has
tr { break-inside: avoid; }. See table page breaks. - Backgrounds don't print until you send
printBackground: true. - Page numbers need a
footerTemplate. It prints at an unreadable size unless it sets its ownfont-size, it overlaps the page unless you set a bottom margin, and your page's CSS doesn't apply to it. See page numbers. - A missing font or image doesn't stop the PDF. Chrome prints a fallback font or a gap, and you find out when someone opens the file.
None of that is hard. It's a few options and CSS rules that you find one bad PDF at a time, and then keep in your code.
PagedPDF has those on by default: rows stay whole, table headers repeat, backgrounds print, "footer": "Page {page} of {pages}" is the whole footer option, and every response lists what failed to load in an X-PagedPDF-Warnings header.
Where Browserless is the better choice
- You need a browser for more than PDFs. Scraping, screenshots, end-to-end tests and logged-in automation all run on the same account.
- PDFs of live pages. It takes a URL, waits for a selector, an event or a function, and can block requests. PagedPDF takes HTML only.
- Full control. You can connect with your own Puppeteer or Playwright code and do anything before printing. PagedPDF has nine options.
- Volume. Above 2,000 PDFs a month it costs less, and its plans run to 500,000 units. PagedPDF's largest plan is 50,000 documents.
- Long jobs. Sessions run from 15 to 60 minutes on paid plans. PagedPDF stops a render at 30 seconds.
- A bigger free tier. 1,000 units a month against PagedPDF's 50 documents.
Where PagedPDF is the better choice
- You only need PDFs, and you'd rather not learn Chrome's printing quirks. The fixes above are already on.
- A bill you can predict. You pay for documents, not seconds, so a slow render costs the same as a fast one. At the limit the API stops with
429 quota_exceededand never charges overage. - 1,000 to 2,000 PDFs a month, where it costs less.
- HTML from your users. Scripts are off unless you turn them on, and each job gets its own browser process that's deleted afterwards.
The same request on each
Browserless, adapted from the example in its documentation:
curl -X POST "https://production-sfo.browserless.io/pdf?token=YOUR_API_TOKEN_HERE" \
-H "Content-Type: application/json" \
-d '{"html": "<h1>Invoice 1042</h1>", "options": {"printBackground": true, "format": "A4"}}' \
-o invoice.pdf
PagedPDF:
curl https://api.pagedpdf.com/v1/pdf \
-H "Authorization: Bearer YOUR_KEY" \
-H "Content-Type: application/json" \
-d '{"html": "<h1>Invoice 1042</h1>", "footer": "Page {page} of {pages}"}' \
-o invoice.pdf