PDF Tools in 2026: How to Merge, Split, Compress, Rotate and Convert PDFs — Privately, in the Browser

A PDF is a container that describes a page as fixed geometry: the text, the images and the drawing instructions are all positioned in absolute coordinates so the page renders identically everywhere, whether or not the reader has the fonts or the original software. That reliability is exactly why it is the format for deliverables — applications, contracts, invoices, prints — and it is also why a PDF resists casual editing. You are not looking at a flow of words you can rewrap; you are looking at a laid-out picture of a page that also happens to carry selectable text.

The practical consequence is the toolkit shape. Everyday PDF jobs are not 'edit this paragraph' — they are structural: combine several files, take pages out, fix the orientation, make the file small enough to upload, or convert it into another format entirely. Each of those is a clean, specific operation, and a set of focused tools does them far faster than opening a heavyweight editor, which is exactly the approach the PDF tools suite takes.

When a PDF is huge, it is almost never the text — plain text is a few kilobytes. The weight lives in the embedded images: scans, phone photos, and pasted screenshots stored at whatever resolution they arrived, often far higher than a screen or an email needs. Compressing a PDF means re-encoding those images at a lower resolution and quality, subsetting the fonts to only the glyphs actually used, and dropping redundant objects. This is why a 15 MB scanned document can shrink to under 2 MB while still reading perfectly on screen.

There is a real trade and it is worth naming. A web or email preset pushes image resolution down hard — great for a form upload or an attachment, wrong for something that will be reprinted, because the pictures go soft. A print-friendly preset keeps higher resolution and saves less. Decide the destination first: the compress PDF tool lets you pick the level rather than guessing, and a page that contains mostly text survives aggressive compression with no visible loss.

These three operations share one mental model — they copy or re-order entire pages without touching what is on them. Merging takes several PDFs and lays their pages into a single document in the order you choose, which is what you reach for when a scanner saved twelve pages as twelve files, or when an application wants 'one PDF' containing ID, transcripts and a letter. The merge PDF tool handles the ordering, and it is lossless page-copying, not re-rendering.

Splitting is the reverse: break one large PDF into separate files, commonly by page ranges, so you can send or archive sections independently. Extracting pages pulls a subset out into a new file — you need only page 7 of a 40-page statement. And deleting pages is just split-and-keep-the-rest. The split PDF and extract PDF pages tools are the two halves of that job; both leave the surviving pages byte-for-byte as they were.

Scanned pages, phone photos and mixed-source documents regularly land sideways or upside-down, and a PDF keeps each page's rotation as a simple attribute. Rotating one, several, or every page is a quick, non-destructive fix that just sets the correct viewing angle — the rotate PDF tool applies 90-degree turns per page or across the whole file. It matters more than it sounds because a portal or print shop will reject an upside-down page the same as a missing one, and nobody wants to flip their monitor to read an application.

The two conversion jobs are opposites and people mix them up constantly. Images into PDF is how you turn a stack of photos or scans into one clean, shareable multi-page document — a receipt bundle, a signed-and-photographed form, a set of phone pictures the recipient's email can display inline. The JPG to PDF tool builds that document, and iPhone users whose camera writes HEIC reach for the HEIC to PDF tool to make the images acceptable anywhere.

PDF into images is the other direction: turn each page into a picture (typically JPG or PNG). You want this when a system only accepts image uploads, when you need a page as a thumbnail, when you are sending one page to a printer on its own, or when you want to view a page without a PDF reader. The PDF to JPG tool rasterises pages at a chosen resolution. If the images are destined for print, the DPI maths in the images-for-printing guide decides how high you should render them.

Two quieter tools cover the 'I just need something out of this PDF without opening an editor' cases. Extracting text pulls the readable text layer into plain text — useful for reuse, translation, accessibility, or searching inside a document that your operating system indexes poorly; the PDF to text tool does this for text-based PDFs (a pure scan has no text layer until it is OCR'd, which is a different job). And PDF metadata is the document's hidden description — title, author, subject, creation date — the file-properties fields rather than page content, which you inspect or clean with the PDF metadata tool before sharing something you would rather not have carry your name.

The strongest reason is the sensitivity of the files themselves. The PDFs people merge and compress are applications, bank statements, contracts, medical forms and government IDs — precisely the documents you should not be happy to ship to a stranger's server to have a job done. Modern PDF engines run inside the browser, so the pages are read, rearranged and re-written on your own device; nothing is transmitted, and you can confirm it the same way you would check any other client-side tool, in the Network tab.

There is an honest limit worth stating so nobody is surprised. A browser-based toolkit is excellent at structural operations — merging, splitting, rotating, compressing, converting, extracting — but it is not a full page-layout editor, and enormous files can run into a tab's memory ceiling. For reflowing paragraph text the way a word processor does, you want a different class of software; for the nine tasks ordinary people actually need on a PDF every week, a local toolset is faster and far safer.

Most real requests reduce to the same handful of steps, and having the order memorised turns a fiddly job into a minute of work.

• Decide the destination first: an email or portal upload wants compression; a reprint wants quality kept.
• To make one document from many, merge in the order you want before compressing, so the size pass runs once over the whole file.
• To send only part of a file, extract or split the exact pages rather than forwarding the entire document.
• Fix orientation with a rotate pass before you submit, not after a rejection.
• When the form will not take a PDF, convert the needed pages to images — and when it will not take images, wrap them into a PDF.
• Check and clear metadata before sharing anything you would rather not have traced back to you.