PDF Tools: What You Actually Need Beyond the Browser

PDF software is bought in a moment of mild panic. Somebody needs two files combined, or a page removed, or a scan that is somehow forty megabytes brought down to something an email server will accept, and the search that follows ends at either a subscription or an advertisement-covered upload form.

Neither is usually the right answer. Most of what people reach for PDF software to do is already possible with what is on the machine, and the small remainder that genuinely is not is a much better guide to whether you should be paying.

The reader is already installed

Every current browser renders PDFs, fills in form fields, adds a drawn signature and saves the result. Firefox does it with pdf.js, Mozilla's Apache-2.0 renderer, which has been built into the browser since version 19 and is also the library to reach for if you ever need to show a PDF inside your own web application.

This is worth stating plainly because a surprising number of people still install a separate reader out of habit, and the separate readers are where the upsell prompts live. If your PDF work is reading, filling and signing, you are finished. There is nothing to buy.

The print dialogue deserves a mention in the same breath. Print to PDF flattens a completed form, drops interactive elements and produces a clean copy of anything the browser can display. It solves more problems than its dull name suggests.

One command-line tool covers the structural work

The next tier of jobs — merge, split, rotate, reorder, repair, strip a password you already know — are structural. They rearrange the file without touching what is drawn on the page, and qpdf does all of them.

• Page selection pulls pages from one or more input files and writes them into a new one, which is both merging and extracting depending on what you ask for.
• Splitting writes each group of pages to its own output file, with a naming pattern you control.
• Decryption produces an unencrypted copy of a file you have the password for — useful when a bank statement arrives locked and you need it in an archive that cannot prompt for a password.
• Linearising rewrites the file so a reader can display the first page before the whole thing has downloaded.
• Checking inspects the structure, encryption and stream encoding and reports what is wrong, which is how you find out whether a file is genuinely damaged or just disliked by one viewer.
• JSON output dumps the structure in a form you can read with ordinary tooling, which turns "why is this file strange" from guesswork into something you can look at.

It installs from every mainstream package manager, runs offline, and is the reason we have not opened an online PDF converter in years. If you are already comfortable in a terminal — and if you are not, our piece on terminal setups that are actually faster is the place to start — this replaces a subscription outright.

Where the free route genuinely stops

Four things sit outside it, and they are worth knowing precisely, because they are the only honest arguments for spending money.

Optical character recognition. A scan is a picture of text. Searching it, copying from it or feeding it to anything else requires recognising the characters first, and that is a different class of software.

Redaction. Drawing a black rectangle over a name does not remove the name. The text stays in the file and comes back out of any extraction tool, and this is how confidential material leaks from documents that somebody believed they had cleaned. Removing the underlying content properly is a specific feature, and if you handle other people's personal data you should be paying for it rather than improvising.

Editing text in place. Changing a sentence inside a finished PDF means reflowing a page that was never designed to reflow. Tools that do it are working against the format. Where you can, fix the original document and export again — the same rule we apply to the files we keep copies of, where the source always outranks the derived artefact.

Tagged, accessible output. If a document has to be usable with a screen reader, it needs a proper structure tree, and adding one after the fact is real work with real tooling behind it.

What the paid option costs

If you cross that line, Acrobat is still the default. UK individual pricing on an annual plan billed monthly runs at £14.99 a month for Standard, £19.97 for Pro and £24.98 for Studio, all including VAT. The month-to-month equivalents are £24.98, £29.96 and £34.98, which is the usual arrangement: the flexible option costs roughly two-thirds more, and most people take it and then keep it for three years.

Our advice on that is the same as it was when we went through what we pay for and what we cancelled: name the specific job before you subscribe. "PDFs" is not a job. "Redacting client names from twenty documents a month" is, and it justifies the licence on its own.

What we would do

Read and sign in the browser. Install qpdf and learn two of its commands. Keep the original document, because every hard PDF problem is easier upstream of the PDF. And pay for exactly one seat of something serious if — and only if — you have a recurring job on that list of four, rather than because a file needed splitting on a Tuesday afternoon.