Three page-level jobs live on this page. Organize edits one PDF on a thumbnail grid: reorder, rotate, delete, or extract pages. Merge combines several PDFs into one, with every page of every file on the same grid. Split divides one PDF into several files. All three copy pages from the source rather than re-rendering them, so text stays selectable and images keep their resolution — and all three run in your browser through pdf-lib, so a contract or medical record is never uploaded.
Organize: the everyday surgery a PDF needs
Documents rarely arrive perfect. A scan comes in with a sideways page, a blank sheet between sections, or pages in the wrong order; a report needs its draft cover removed; you want just the signature page out of a long contract. Doing these visually, with thumbnails you can drag, is faster and less error-prone than juggling page numbers in a dialog box.
- Reorder — drag thumbnails (or long-press on touch screens) to fix the sequence.
- Rotate — per page or all at once, written into the file rather than just shown.
- Delete — drop blank, duplicate or confidential pages before sharing.
- Extract — tick pages and save only those as a new PDF.
Keyboard users can do the same without a mouse: arrow keys move between pages, R rotates, Delete removes, Space selects, and Shift extends a selection.
Merge: one file beats a folder of attachments
Sending five PDFs asks the recipient to download, order and track them; one merged PDF hands them a finished document — a job application, scanned contract pages stitched into one agreement, a month of invoices for an expense claim. Because the result is read front to back, the order you arrange pages in is the reading experience. Add the files, drag pages into sequence, prune blank scanner pages, and check the joins between documents, where mistakes hide. Mixed page sizes are fine; each page keeps its own dimensions.
What a merge silently throws away
A PDF is more than its pages, and the parts that don’t survive a merge are the parts nobody warns you about.
Form fields collide by name. PDF identifies an interactive field by its fully-qualified name, not by which document it came from. Merge two filled copies of the same template and both copies of applicant_name compete for one value. If you’re combining filled forms, flatten each one first.
Bookmarks vanish. The outline tree lives in the document catalogue, not in the pages, so a page-level copy leaves it behind.
Page labels are lost. If a document numbered its front matter i, ii, iii, that mapping is document-level metadata; after merging, a viewer shows plain positions.
Encrypted files must be decrypted first, which needs the password that opens them.
None of this is a reason to avoid merging. It’s a reason to merge last, once each document is final.
Split: when one PDF should be several
Splitting lifts out the pages a task needs: one signed page filed separately, a client’s section of a combined report, a scanned book broken into chapters, a batch scan where several documents landed in one file.
- Single pages — one PDF per page; right for batch scans.
- Custom range —
2-4, 9makes one file per range; right for pulling sections out of a report. - Every N pages — equal chunks, such as ten-page sections of a 100-page manual.
- Extract — tick pages on a grid and save them together as one PDF.
A preview line tells you how many files you’ll get before you split. Several results can be downloaded one by one or together as a ZIP.
Checking that nothing is uploaded
You don’t have to take the privacy claim on faith: open DevTools, switch to the Network panel, and merge or split a file — no request carries your document. Or disconnect from the network entirely; all three modes still work, because there was never a round trip to lose.