Skip to content
Cheeky Tools

Split a PDF

File ToolsExtract the pages you want, or burst a PDF into one file per page — never uploaded.

Your files never leave this device. Everything here runs in your browser — your files are never uploaded, stored or seen by us. Open your network tab and check.

Single pages, ranges, or 8- for everything from page 8.

Add a PDF. Pages are copied out without re-encoding, so nothing loses quality.

How this is calculated

Extract:  new PDF = the pages you selected, in the order you wrote them
Burst:    one new PDF per page, numbered in order

Page selection syntax
  1          a single page
  1-3        an inclusive range, pages 1, 2 and 3
  8-         page 8 to the end
  1-3, 7     combine them with commas

Every output is a rebuilt document with the source metadata removed.
Nothing is uploaded. The split runs in this browser tab.

How it works

Splitting a PDF copies whole page objects — along with the fonts and images those pages reference — into a new document. Nothing is re-encoded, so text stays selectable and images keep the quality they already had. That is why a single page pulled from a large document is usually small, but not always: a page carrying a big embedded font or a high-resolution scan brings that weight with it.

There are two things people mean by split. Extracting is pulling a range out into one new file — the pages you need from a longer report, or the signed page from a contract. Bursting is turning a document into one file per page, which is what you want when each page is a separate record and someone downstream needs them individually.

The page selection is written the way you would say it: 1-3, 7, 10-12. A trailing hyphen means to the end, so 8- takes everything from page 8 onward. Order is preserved as typed, so 5, 1 really does put page 5 first — occasionally useful, and never assumed to be a mistake.

What the parser will not do is guess. A page number past the end of the document is an error, not something to clamp to the last page. A range written backwards is an error, not something to quietly reverse. Two numbers with a missing comma are an error, not a range. Every one of those would produce a file that looks correct and is not, which is the worst outcome a split tool can have.

Worked example

Pulling the signed pages out of a 40-page agreement

  1. Original 40 pages
  2. Selection 38-40
  3. Parsed as pages 38, 39, 40
  4. ---------
  5. Output 1 file, 3 pages
  6. Selection 1-3, 38- would give pages 1, 2, 3, 38, 39, 40

A 3-page PDF, with the original's title and author stripped

How to use it

  1. Choose extract to pull pages into one new file, or one file per page to burst the document.
  2. Drop the PDF in. The page count is read straight away so you know what you are working with.
  3. Type the pages you want: 1-3, 7, 10-12. The hint under the field tells you how many pages that comes to before you commit.
  4. Extract, then download. Nothing is uploaded at any point.
  5. To reorder pages rather than remove them, extract them in the order you want — 3, 1, 2 is a valid selection.

What this tool does not do

  • Password-protected PDFs cannot be split. They will open for reading but refuse to have pages copied, and you will see a message saying so.
  • Bursting gives one download button per page rather than a zip. A hundred-page document produces a hundred buttons, which is honest but tedious — extract a range instead if that is what you actually need.
  • Bookmarks, form fields, annotations and digital signatures are not carried into the output. Pages, text, fonts and images are.
  • Output size is not proportional to page count. A single page can carry a large embedded font or image, so a one-page extract from a small file is sometimes almost as big as the original.
  • It runs in your browser, so very large documents are limited by your device's memory rather than by a server quota.

Frequently asked questions

Are my files uploaded anywhere?

No. The split runs in your browser using pdf-lib, and no network request carries your file. You can check: open your browser's network tab before splitting and watch that none of them does — the only requests you will see after the page loads are small ones recording that the tool was used. Most online split tools post your document to a server first.

How do I write the page selection?

Single pages and ranges, separated by commas: 1-3, 7, 10-12. A trailing hyphen means to the end, so 8- takes page 8 onward. The hint under the field shows how many pages your selection comes to before you run it.

What happens if I type a page that does not exist?

You get an error naming the problem, not a file. If your document has 20 pages and you ask for 25, the tool says so rather than quietly giving you page 20. Clamping a typo to the nearest valid page is how you end up with a document that looks right and is wrong.

Can I reorder pages with this?

Yes, within an extract. The selection is used in the order you write it, so 3, 1, 2 produces a file with those pages in that order. You can also repeat a page — 1, 5, 1 is allowed, because sometimes that is genuinely what you want.

Why do I get one download per page instead of a zip?

Because building a zip would mean adding a compression library for one feature, and this site keeps its dependencies to what the tools actually need. For a handful of pages the buttons are fine; for a long document, extracting the range you want is usually the better move anyway.

Does the output keep bookmarks and form fields?

No. Pages, text, fonts and images are copied across. Bookmarks, interactive form fields, annotations and digital signatures are not — and a signature would be invalidated by the split in any case.

Why is my extracted page almost as large as the original?

Because a page brings its resources with it. If the document embeds a large font subset or a high-resolution scan that your page uses, that weight comes along. Run the result through the PDF compressor if size matters more than fidelity.

Can I split a scanned PDF?

Yes. Scanned pages are images inside a PDF container, and they copy across exactly as they are. The output will not have selectable text if the original did not, because splitting does not run OCR.

Related tools

Spotted something wrong?
tell us

The figure you got and the figure you expected, if you have both. A source helps.

Only so we can reply. Leave it blank and we still read it.

— every formula on this page is published so it can be checked.

Last updated 7 September 2026.