Do I have to use the sheets made here?

No. Any page with a QR code can be a separator: leave “Code must start with” blank and every page with a code counts, or keep a prefix (SEP: by default) so only matching codes do — then documents that carry their own QR codes, like invoices, aren’t split by mistake. You can also skip QR codes and use plain blank sheets, or sheets with a fixed printed word (this needs a text layer).

What about blank backs when scanning double-sided?

Turn on both QR and blank page detection: the blank back of a separator sheet is treated as a separator as well, and since there’s nothing between consecutive separators no empty file is created. If your documents contain intentional blank pages, turn blank detection off and rely on QR codes.

How do I fix a wrong detection?

Click the page thumbnail to cycle: normal → separator → start a new file here (the page stays in the new file) → normal. The split plan below updates immediately, and “Undo manual changes” restores the automatic result.

Is my scan uploaded, and does splitting reduce quality?

Nothing is uploaded — rendering, detection and splitting all happen in your browser. Splitting copies the original page objects without recompressing images, so the output looks exactly like the input; rendering is only used to read codes and spot blank pages.

🚩

🚩Split Scanned PDF by QR Separator Sheets or Blank Pages

Scanning dozens of contracts or receipts in one go is quick, but splitting the resulting PDF afterwards is not. Start by making separator sheets: enter one name per line (such as “Contract-Smith”) and get an A4 page for each with a large vector QR code and the name printed on it; the QR code holds your prefix plus the name. Put a sheet in front of each document, scan the whole stack, and drop the PDF in. Every page is rendered and checked: QR codes and barcodes are read (more barcode types where the browser supports them) and a page counts as a separator when its code starts with your prefix. Blank pages — judged by how much of the page has content, ignoring edge shadows — or pages carrying a marker text such as “SEPARATOR” can act as separators too. Separator pages can be removed, kept as the first page of the next file, or kept as the last page of the previous one. Results appear as thumbnails showing each separator and why it was detected; click any page to cycle between normal, separator and “start a new file here” to fix misses or false hits. The plan lists the pages of every output file, named after the separator’s QR content by default and editable one by one, and everything downloads as a ZIP. Pages are copied as they are, not recompressed, so quality is unchanged. It all runs locally in your browser; nothing is uploaded.

Free · No sign-up · Runs in your browserLast updated: 2026-10-26

Loading tool...

How to use Split PDF by Separator Sheets

1Under “Make separator sheets”, enter one name per line, print the QR sheets and place one in front of each document when scanning
2Drop in the scanned PDF and choose how separators are detected (QR, blank page, text) and what happens to them
3Check the thumbnails, click to fix any split points, adjust the names and download

FAQ

Do I have to use the sheets made here?: No. Any page with a QR code can be a separator: leave “Code must start with” blank and every page with a code counts, or keep a prefix (SEP: by default) so only matching codes do — then documents that carry their own QR codes, like invoices, aren’t split by mistake. You can also skip QR codes and use plain blank sheets, or sheets with a fixed printed word (this needs a text layer).
What about blank backs when scanning double-sided?: Turn on both QR and blank page detection: the blank back of a separator sheet is treated as a separator as well, and since there’s nothing between consecutive separators no empty file is created. If your documents contain intentional blank pages, turn blank detection off and rely on QR codes.
How do I fix a wrong detection?: Click the page thumbnail to cycle: normal → separator → start a new file here (the page stays in the new file) → normal. The split plan below updates immediately, and “Undo manual changes” restores the automatic result.
Is my scan uploaded, and does splitting reduce quality?: Nothing is uploaded — rendering, detection and splitting all happen in your browser. Splitting copies the original page objects without recompressing images, so the output looks exactly like the input; rendering is only used to read codes and spot blank pages.