Split one PDF into many, automatically

You fed ninety invoices through a scanner and got back one three-hundred-page PDF. Ordinary split wants you to tell it where the boundaries are, which means finding all ninety of them by hand. This finds them.

or drop a file anywhere on this panel

opened locally · 0 bytes sent

How it works

  1. 01Open the PDF.
  2. 02Choose what marks the start of each document: a bookmark, a blank page, a phrase, or a fixed number of pages.
  3. 03Check the list of parts the panel shows.
  4. 04Download — you get a zip of separate PDFs.

Four rules, one of which will fit

Split at every bookmark, and each chapter becomes a file named after its heading. Split at blank pages, the separator sheets people feed between documents. Split wherever a phrase appears — "Invoice number", "Dear", a case reference. Or simply every N pages, when the documents are all the same length.

Every page ends up somewhere

The parts together contain every page of the original, in order, with none duplicated and none lost. The one exception is blank separator pages, which are discarded by default because they were never part of any document — and you can keep them if you would rather.

You see the split before it happens

The panel lists every part with its page range and, where it can, its name, before you download anything. A rule that finds no boundaries says so plainly rather than quietly handing you a zip containing one file identical to what you started with.

What "blank" means here, exactly

A blank page is one with nothing drawn on it. A scanned blank sheet is not that — it is a photograph of white paper, which as far as the file is concerned is a full-page image. So blank-page splitting works on documents assembled digitally, and not on raw scans. The panel says this when it finds nothing, instead of leaving you to guess.

Files named after their contents

Splitting at bookmarks names each file after the bookmark, with a zero-padded number in front so your file manager sorts them in document order rather than 1, 10, 11, 2. Characters that are illegal in a filename are stripped rather than producing a file that will not save.

Questions

Why is the download a zip?
Because the output is many files. The zip is built in your browser, uncompressed — the PDFs inside are already compressed, and zipping them again would spend time to save nothing.
How many parts can it make?
Up to five hundred. Past that the rule is almost certainly matching more than you meant, and the panel says so rather than building a zip nobody wants.
Can I split by page number ranges instead?
Yes — that is the ordinary Split tool, which takes ranges like "1-4, 9, 12-". This one is for when you do not know the ranges.