Every real PDF-splitting job reduces to one of three shapes: cut here-to-here (ranges), cut every N pages (chunks), or cut everywhere (burst). Once you know which shape you need, the tool — or the one-line command — follows immediately. Here's the syntax, the arithmetic that catches people out (remainders), and what happens to quality and file size when you cut a document up.
Each output file is a fresh, valid PDF containing byte-for-byte copies of the pages you selected, along with the fonts and images those pages reference. Nothing is recompressed: text stays selectable, scans stay scans, page sizes pass through unchanged. One side effect worth knowing — each small file carries its own copy of shared resources, so splitting one 10 MB document into ten pieces can total slightly more than 10 MB. That's duplicated fonts, not degraded quality.
Ranges are the scalpel. Type comma-separated selections in a human format:
1-3 — pages 1, 2, and 3 (inclusive on both ends)5 — just page 51-3, 5, 8-10 — three selections, three output files (3 + 1 + 3 pages)9- — page 9 through the last page-4 — page 1 through 4On a 12-page document, 1-3, 5, 8-10 yields three PDFs: one with pages 1–3, one with page 5, one with pages 8–10. Pages 4, 6, 7, 11, and 12 are simply not included — that's the point. Range splits are how you pull a chapter, redact by omission (send the version minus the pages you don't want seen), or assemble the pieces of a custom packet.
The tool cuts at fixed boundaries of N pages and puts the remainder in the last file. The arithmetic: a 12-page PDF split every 5 pages gives ceil(12 ÷ 5) = 3 files — pages 1–5, 6–10, and 11–12, i.e. 5, 5, and 2 pages, summing back to exactly 12. Every page lands in exactly one output; nothing is duplicated or dropped. Even divisions are cleaner: 12 split every 4 gives three equal 4-page files.
| Pages | Split every | Resulting files | Pages per file |
|---|---|---|---|
| 12 | 5 | 3 | 5, 5, 2 |
| 12 | 4 | 3 | 4, 4, 4 |
| 10 | 3 | 4 | 3, 3, 3, 1 |
| 25 | 10 | 3 | 10, 10, 5 |
Chunk mode exists for hard constraints: email attachment ceilings, upload limits on portals that cap page counts, and print batches that need to move in fives or tens.
Burst — one file per page — is chunk mode with N = 1, and it's what you want when pages are processed individually: per-page review, feeding a fax-style workflow, dropping scanned pages into a system that takes one at a time, or building a page thumbnail set. A 6-page document becomes 6 PDFs, named file-1.pdf through file-6.pdf.
Two tools cover it. With qpdf, a range split is qpdf in.pdf --pages . 1-3 -- part1.pdf (the . means "from the input file"; the range is 1-based and inclusive), and every-page splitting is qpdf --split-pages in.pdf. With pdftk: pdftk in.pdf cat 1-3 5 output parts.pdf concatenates selected ranges into one file, pdftk in.pdf cat 1-3 output part1.pdf writes them separately, and pdftk in.pdf burst output page_%02d.pdf is burst mode. The browser splitter is the same operation without a terminal — and without your document leaving the machine.
Splitting makes many files; extraction makes one file from a selection. If you need pages 2, 5–7 as a single new PDF, either use the CLI above (qpdf in.pdf --pages . 2,5-7 -- extracted.pdf) or do the two-step in the browser: split those ranges out, then recombine them with the PDF Merge tool. The split→merge pair also solves reordering: cut a document into its new sections, then merge them back in the order you want.
Ranges, chunks, or burst — in your browser, no upload, no watermark, plan preview before you commit.
Split a PDF →Pick the shape first: ranges for selections, chunks for equal cuts, burst for everything. Know that the last chunk holds the remainder (12 every 5 = 5+5+2), that quality never changes because pages are copied not re-rendered, and that encrypted files must be unlocked before any tool will touch them. For one-off jobs the browser tool is fastest; for pipelines, qpdf and pdftk are the same math in a terminal.
Type comma-separated ranges — 1-3, 5, 8-10 — and each range becomes its own PDF. A 10-page document split as 1-3, 5, 8-10 produces three files with 3, 1, and 3 pages. Ranges are inclusive on both ends; a bare number is a single page; 9- means page 9 through the end; -4 means pages 1 through 4.
Use fixed chunks of N pages. The tool cuts at multiples of N, and the last file holds the remainder: a 12-page PDF split every 5 pages gives three files of pages 1-5, 6-10, and 11-12 — that's 5, 5, and 2 pages. Page counts that divide evenly (12 split every 4) produce equal files with no remainder.
Split with the single page number as the range — type 7 for page 7 — and you get a one-page PDF. In a terminal: qpdf in.pdf --pages . 7 -- page7.pdf. For every page as its own file, use burst mode (pdftk in.pdf burst is the CLI version).
Only after unlocking it. Encrypted PDFs can't be read by any splitter — browser or desktop — while locked. Open the file with its password in a viewer and re-save an unlocked copy (or print-to-PDF), then split that. Owner-password restrictions are the usual reason a file refuses to load.