Split Scanned PDF

One scan in, separate named documents out

Sheet-feed scanners are wonderful right up until the moment they hand you the result: a single PDF with a month of receipts, three signed contracts, and a couple of letters all welded together. Every document management guide tells you to keep files separate and named. Nobody tells you how to get there from a 60-page blob without an afternoon of manual splitting. This tool is the missing step: upload the blob, download a ZIP where each document is its own PDF with a sensible name.

The blank-sheet trick

The default mode looks for blank pages and treats them as separators. That's not an arbitrary choice; it's the oldest trick in high-volume scanning. Before feeding the stack, slide one blank sheet of paper between each document. The scanner dutifully scans the blanks along with everything else, and this tool finds them, cuts the file at each one, and drops the blank pages from the output. Your five documents come out as five files, and no blank page survives into any of them.

Detection is tolerant of reality. A scanned blank sheet is never digitally empty; there's paper texture, a shadow at the edge, the odd speck. A page counts as blank when it has no text layer and almost no ink, which catches real-world blank sheets while leaving genuinely sparse pages, a title page with three words on it, safely on the document side of the line.

Where the file names come from

Each output file is named from the first meaningful line of its opening page. If the scan already carries a text layer, that text is used directly; if it's a pure image scan, the first page is OCRed just for naming. A document opening with "Rental Agreement" becomes 02-Rental-Agreement.pdf. The numbered prefix preserves the scan order, so the ZIP lists your documents in the sequence they went through the scanner, and duplicate names get a numeric suffix instead of overwriting each other.

An honest caveat: the name is only as good as the first line. A document that opens with a logo and no readable text falls back to a plain document-03 style name, and a cover sheet that starts with a date gives you a date-named file. It's a suggestion engine, not a mind reader, and renaming the occasional file still beats naming all of them.

Every-N mode for scans without separators

Already sitting on a scan with no blank sheets in it? If your documents have a fixed length, two-page delivery notes, four-page report cards, switch to every-N mode and the file is cut at fixed intervals, with the same automatic naming applied to each piece. For irregular documents with no separators and no fixed length, no tool can guess the boundaries reliably, and the honest advice is to rescan with blank sheets between documents. Thirty seconds of paper shuffling saves the guessing entirely.

And when you know the exact page numbers you want, our regular Split PDF tool cuts by explicit ranges instead of detection.

A real example

A bookkeeper closes out the month with thirty client receipts. She stacks them with blank sheets between, runs the pile through the feeder once, and uploads the resulting PDF here. A minute later she has a ZIP of thirty files with names like 07-Office-Depot-Receipt.pdf, in scan order, ready to file into the accounting system. The alternative was thirty rounds of open, find the page range, extract, name, save. That's the whole pitch, really: the scanner already did the hard part, and this finishes the job.

Split too eagerly? You can always merge them back together in whatever order you like.

Honest limitations

• Duplex scans of single-sided documents produce a blank back for every sheet, which blank-page mode reads as a separator after every page. Use every-N mode (N of 2) for those, or scan single-sided.
• Name suggestion reads Latin-script text best; names for scans in other scripts may fall back to numbered files.
• Splitting doesn't make the documents searchable; the pages come out exactly as scanned.

To make the separated documents searchable too, run each one through our OCR PDF tool afterward; its review-copy option even highlights the words it wasn't sure about.

• The per-file upload limit is shown on the tool, and very long scans simply take a little longer to analyze.

Privacy and file handling

Batch scans are usually paperwork: receipts, agreements, medical letters, the stuff of filing cabinets. The handling is accordingly strict. Your upload travels encrypted, is processed by an isolated routine, and is deleted from the server the moment the job completes, along with every intermediate page image the analysis created. The ZIP you download is the only copy that leaves the process.