Split PDF
Write ranges like "1-3,5,7-9"; each range produces one file.
Drag in one PDF, or Choose file
Processed locally · No upload"1-3" means pages 1 through 3, not indexes. Ranges beyond the total page count are ignored with a notice.
When to use it
- A 30-page contract needs only pages 1–3 (the quote) and the signature page sent to the other party.
- Every page of a scanned or exported PDF is a separate invoice or certificate, and you need one file per page for filing.
When not to
- You just want to jump to certain pages while reading — no new file is needed; type the page number in your PDF viewer.
- You want automatic splitting by bookmarks or chapters with the original outline intact — this tool cuts by page ranges only and doesn't read the outline.
Splitting runs locally through pdf-lib — the PDF is never uploaded and is gone once you close the tab. Page numbers start at 1, and ranges beyond the last page are clipped or ignored. Password-protected files can't be read. Splitting page by page creates one new PDF per page in memory, so very long documents can run out of memory on a phone, and bookmarks or form fields aren't preserved.
Frequently asked
Is the PDF uploaded when I split it?
No. pdf-lib copies the pages in your browser and builds the new files there; both the original and the output stay on your device. The first split loads a ~500 KB pdf-lib script, which is reused for the rest of the session.
How do I write page ranges, and what happens if they're out of range?
Numbering starts at 1. Separate segments with commas, for example 1-3,5,7-9, and each segment becomes one file. Anything past the last page is clipped: on a 10-page PDF, 8-20 outputs pages 8–10 only. A segment with no valid numbers is rejected.
Can I split a several-hundred-page PDF?
There's no hard page limit; it depends on memory. A few hundred ordinary pages usually work, but splitting every page creates that many PDFs at once, which can slow a phone down or fail. Outputs with more than one file are packed into split.zip automatically, and using ranges instead of per-page mode produces far fewer files.