Your paperwork is split across three PDFs and the form wants one file. Or the opposite: a 30-page contract, and you only need to send the two signature pages. Both are routine, and neither is hard — but the tool you pick changes what you end up with more than you would expect.
The question is whether anything gets redrawn
There are two broad ways to work with a PDF. One copies the page objects as they are. The other renders each page to an image and pastes that image into a new PDF. They look the same on screen and are completely different documents underneath.
| Copying page objects | Re-rendering as images | |
|---|---|---|
| Text | Stays text — searchable, selectable, translatable | Becomes a picture — neither searchable nor selectable |
| Quality | Identical to the original | Depends on the render resolution |
| File size | Roughly the sum of the inputs | Can grow or shrink dramatically |
| Fonts | Preserved | The concept no longer applies |
So for merging and splitting you want the copying approach. You can tell which one a tool used by trying to select text in the result. If nothing highlights, your document has become a photo album.
Merging — order is the document
In a PDF, page order is not a sorting preference; it is the document itself. A good merge tool therefore follows the order you added the files in. Tools that silently sort by filename are where 'chapter1.pdf, chapter10.pdf, chapter2.pdf' comes from. Check the order before you start.
Expect the result to be about the size of the inputs added together. Documents that share fonts may come out slightly smaller as duplicated font resources are consolidated. What merging will not do is make a large file meaningfully smaller — that is compression's job, not merging's.
Splitting — 'extract' and 'one file per page' are different jobs
Splitting covers two distinct tasks: pulling two signature pages out of a contract into a single PDF, and cutting a 30-page scan into 30 separate files. The first is extraction, the second is per-page splitting. Pick the wrong one and you will find yourself merging thirty files back together.
- Page ranges are usually written like '1-3, 7, 10-12' — commas separate entries, hyphens mark spans.
- Those numbers mean the nth sheet in the file, not the number printed on the page. A cover and a table of contents will push the two out of step.
- When the result is many files, downloading them as one ZIP is far less tedious.
What can disappear
Even when page objects are copied faithfully, anything that lives above the page level may not come along. Bookmarks and form fields are the usual casualties. An outline is a structure spanning the whole document, so a newly created document has none unless it is rebuilt. Form fields likewise depend on document-level definitions and may stop working after a merge or split.
- Bookmarks and outlines — assume they are gone.
- Form fields — may come out empty or non-interactive. If a form is already filled in, flatten it to a print-style copy before merging.
- Digital signatures — any change to the document breaks them. As a rule, do not merge or split a signed PDF.
- Password-protected PDFs — must be unlocked first. Tools simply refuse a file they cannot read.
Handling paperwork without uploading it
The things people keep as PDFs tend to be contracts, ID scans and payslips — documents you would rather not hand to a stranger. Yet most online PDF tools upload the file to a server before touching it. Switching merges and splits entirely inside your browser, so the file never leaves your device. It handles up to 20 files and 100MB per merge, and up to 100 pages per split.