Split PDF

Split PDF: click between pages to cut, and a new file starts right there

To split a PDF is to turn one PDF into several files, and to extract pages is to pick only the pages you need and save them as a new PDF. This tool does both on one screen. Click the scissors between page thumbnails and a new file starts at that spot; click a page and it drops out of the result. How many files you will get, and which pages go into each, shows up in the ribbon and file chips above before you save anything.

For example, splitting the sample PDF (12 pages, 204 KB) by bookmark gives six PDFs: cover, contents, chapters 1 to 3 and the appendix, each named after its chapter. Pages are copied from the original as they are, so text does not turn into pictures and page size and rotation stay the same. The PDF you choose is never sent to a server; it is processed inside the browser you are using.

When to split a PDF

Sending only the pages that were asked for: if you need to hand in a few pages from a report or a contract bundle, keep those and leave out the rest. Pages with personal details, or anything the other side does not need, do not have to go along.
Getting under an email attachment limit: a personal Gmail account limits attachments to 25MB in total, and anything over that is added as a Google Drive link instead. Splitting a large PDF by chapter lets you fit each part into one email. The total of the split files can be larger than the original, though, so read ‘Size of the split files’ below first.
Separating a batch scan into one file per case: if you scanned several application forms in one go, split every N pages (the page count of one form), or click the scissors wherever a new form begins.
Splitting a book, textbook or report by chapter: for a PDF with bookmarks, press [Each bookmark] and it is cut before every page where a chapter starts, with the chapter name in the file name.
Tidying a double-sided scan: you can keep only the odd pages or only the even pages and save them as one file.

How to use it

Drop a PDF on the work area or press [Choose PDF]. If it is your first time, press [Open sample PDF] to open a made-up 12-page report.
Decide where to cut. Press [Every page], [N pages each] or [Each bookmark], or click the scissors between pages on the page board yourself.
Click the pages you do not want in the result. [Odd pages], [Even pages] and [Invert] change many pages at once.
Check the number of files and the pages of each one in the ribbon and file chips above. The same thing appears as text in the [Type ranges] box, and if you edit that box the page board follows.
Press [Save n as ZIP] to get every file in one ZIP. If you need just one file, press its chip. When only one file will be made, the button changes to [Save PDF].

Reading the page board and the ribbon

The strip under each thumbnail shows which file that page goes into. Pages in the same file share a strip that runs on to the scissors; where the file changes, the strip breaks and a dotted line and scissors appear. Neighbouring files alternate a dark and a light colour, and the first page of each file carries its file number.

A page with hatching and an ✕ is left out of the result. If a page that is left out sits in the middle of a file, a thin dotted strip runs beneath it, meaning the pages on either side are in the same file.
The small figure beside a page number is the printed page number set in the document. When you write ranges, use the bold screen number.
The ribbon lays the whole document out as equal-width pages in one bar. Each file has its own colour and left-out pages show as hatching. In a long document where the page board scrolls, the range you can see now is outlined, and clicking the ribbon moves the board to that page.
Keyboard: on the page board, move between pages and scissors with the arrow keys and press Space or Enter to click. Hold Shift and click another page to keep or leave out everything in between.

Ways to split, compared

Method	What it does	Result for the sample PDF (12 pages)	Good for
No split	Keeps the pages you kept in one file, with no cuts	Keeping only odd pages gives one 6-page PDF	Pulling just the pages you need into one file
Every page	Each kept page on its own	12 PDFs	Submitting or uploading page by page
N pages each	Kept pages in groups of N, with only the remainder in the last file	5 pages each gives 5, 5 and 2 pages: 3 PDFs	Scanned batches where each case has the same page count
Each bookmark	Cuts before the page where each top-level bookmark starts	1, 1, 3, 3, 2 and 2 pages: 6 PDFs	Reports and books to split by chapter
Scissors between pages	Cuts at every spot you click	Whatever you click	Irregular boundaries
Type ranges	Splits as written and leaves out pages you do not write	1-2; 3-8; 11-12 gives 3 PDFs (pages 9 and 10 left out)	When you already know the page numbers

N pages each does not count pages that are left out. In the sample PDF, leaving out the cover (page 1) and splitting 5 pages each gives three files: pages 2 to 6, 7 to 11, and 12. A split button looks pressed only when the current state of the page board matches that method, so clicking one more pair of scissors can leave no button pressed.

How to write ranges

In the [Type ranges] box, a comma (,) adds pages to the same file, and a semicolon (;) or a line break starts a new file. 3-8 includes both page 3 and page 8, 9- means page 9 to the last page, and -3 means pages 1 to 3. A tilde (3~8) or a long dash works in place of the hyphen, and full-width digits are accepted too.

What you type (12-page document)	Result
3	One PDF with page 3 only
2, 4-6	One PDF with pages 2, 4, 5 and 6
1-2; 3-8; 9-	3 PDFs (2, 6 and 4 pages)
6, 2, 2, 4-5	One PDF with pages 2, 4, 5 and 6 (pages inside a file keep the original order, and a repeated page counts once)
9-12; 1-2	The file with pages 1 to 2 is number 1 and the file with pages 9 to 12 is number 2 (file numbers follow the order of first pages)

If you write any of the following, the page board stays as it is and the reason appears under the box: a page the document does not have, such as 0 or 13; a range written backwards, such as 5-3; an empty spot between commas, as in 1,,2; a value that is not a whole number, such as 1.5; one page in two files, as in 1-3; 3-5; or files whose pages cross over, as in 1, 5; 2-4. This tool cuts in the original order, so it does not put the same page into several files or reorder pages.

Printed page numbers and screen numbers

In the PDF standard (ISO 32000-1 12.4.2) a page has a number fixed by its order in the document, and separately the document can define page labels to show on screen or in print. The usual example is front matter numbered with Roman numerals and the main text with Arabic ones. The ranges and the page board in this tool use the screen number, which counts the first page as 1.

In the sample PDF the cover and contents are i and ii, and numbering starts at 1 on the first page of chapter 1. So the first page of chapter 1 is screen number 3, and the page board shows both numbers, like ‘3 1’. To pull out chapter 1 you write 3-5, not 1-3. Each split file counts from 1 again, and the numbers printed inside the pages do not change.

Splitting by chapter with bookmarks

The bookmarks of a PDF (its document outline) are a table of contents holding chapters and sections as a tree. [Each bookmark] cuts before every page that a top-level bookmark points to. If there is only one top-level bookmark and several items under it, those items are treated as the chapters. Bookmarks that point to the same page count as one, and bookmarks that open a web address are not used as split points.

The sample PDF has six top-level bookmarks (Cover, Contents, Ch. 1 Usage, Ch. 2 Programs, Ch. 3 Next Year and Appendix Tables) and section bookmarks under chapters 1 and 2. Section bookmarks are not top level, so they do not become split points. When a file holds exactly one whole chapter, the chapter name goes into the file name (for example sample-library-report_03_Ch. 1 Usage.pdf), and when it holds only part of a chapter, the page range is used (for example _03_p3-4.pdf).

What stays in the result and what does not

The tool does not redraw pages. It copies the content objects of the original pages into a new PDF. That settles what stays and what goes, as follows.

Stays: the text, pictures and shapes on the page, page size and rotation, links to web addresses, and annotations such as highlights and notes. Text remains text, so you can still search and copy it.
Goes: links that jump to another page inside the document, such as contents links; replies attached to a note on another page; bookmarks; the printed page number setup; document properties (title and author); the tag structure used by screen readers; and actions that run when a page opens, plus article-flow information. An item that points to another page could drag a left-out page into the result file, so these are detached before copying.
PDFs that are not opened: password-protected PDFs, and PDFs with fill-in fields (forms) or signature fields, because splitting could lose what was entered. If the form is already filled in, choose Print in a PDF viewer and then ‘Save as PDF’ to make a copy with the entries built into the page content, and split that.

Size of the split files

When a page is copied, the fonts and images it uses are copied with it. If several pages share one font, every split file carries that font, so the files add up to more than the original. The sample PDF (204 KB) has 12 pages that share one font. Split by bookmark into 6 files, the total is 944 KB; split every page into 12 files, it is 1.8 MB, and even the smallest single-page file is 152 KB.

Material used on only one page, such as a photo, goes only into the file that contains that page. After saving, each file chip shows its size, and the status line shows the total next to the original size. If you are splitting because of an email limit, check these sizes, and if one file is still large, shrink its photos with Compress PDF.

Common mistakes

Writing the printed number in a range: use the bold screen number on the page board, not the figure printed at the foot of the page.
Swapping the comma and the semicolon: 1-2, 5-6 is one file of four pages, while 1-2; 5-6 is two files of two pages.
Counting left-out pages in N pages each: left-out pages are not counted, so after you leave pages out, N pages each splits the pages you kept.
Trying to put one page into several files: this tool puts a page into one file only. If every file needs the cover, save once per file with a range that includes the cover.
Looking for the files without unzipping: several files are saved as one ZIP. If a ZIP is awkward on your phone, press the file chips one at a time to get each as a PDF.
Changing the split after saving: files you have already downloaded do not change. Save again after making the change.

Limits and cautions

One PDF is opened at a time, up to 100 MB and 1,000 pages. The files made at one time add up to at most 200 MB.
Pages are not redrawn, so the tool does not shrink file size, do text recognition (OCR), cut a page in two side to side, or reorder pages.
Splitting a signed PDF would make its signature fail verification, so this tool does not open it. Keep the signed original separately.
Some PDFs cannot be previewed. Even then, you can split by choosing pages by number.
Splitting a large document into many files uses a lot of browser memory. If it stalls on a phone, reduce the number of files or split on a computer. This tool was tested in Chrome, and speed and memory use in other browsers were not measured.

Date checked The standard, the help page and the example figures were checked on 2026-10-21.

Sources

01ISO 32000-1:2008 Document management — Portable document format (public copy from Adobe)12.3.2 Destinations, 12.3.3 Document outline (bookmarks), 12.4.2 Page labels (the page order number and the printed number can differ), Table 164 annotation P entry, 12.5.6.5 Link annotations
02pdf-lib API — PDFDocument.copyPagesThe function that copies pages from one document into a new one. This tool builds files with version 1.17.1 (MIT license)
03pdf.js 3.11.174 — src/display/api.jsgetOutline (bookmarks), getDestination, getPageIndex (the page a bookmark points to) and getPageLabels (printed page numbers). Thumbnails are drawn by this engine too (Apache-2.0 license)
04PKWARE — .ZIP File Format Specification (APPNOTE.TXT)Compression method 0 = stored without compression; general purpose bit 11 = file names recorded in UTF-8 (non-English file names)
05Gmail Help — Attach files to your emailAttachment limit of 25MB for a personal Gmail account; anything over is added as a Google Drive link

Frequently asked questions

Is my PDF uploaded to a server?

No. The PDF you choose is read and split inside the browser you are using. Only the tool files needed to run the page are downloaded, and the contents of your PDF are not sent anywhere. Choosing another PDF or closing the page clears your work.

How do I keep only the pages I want in one file?

Press [No split], click the pages you do not need on the page board to leave them out, then press [Save PDF]. If you know the page numbers, you can also type 2, 4-6 in [Type ranges]. That makes one PDF containing only pages 2, 4, 5 and 6.

Can I save every page as its own file?

Yes. Press [Every page] and each kept page becomes its own file. A 12-page document gives 12 PDFs saved in one ZIP, and pressing a file chip downloads just that page as a PDF.

How do I split a PDF by chapter?

If the PDF has bookmarks, press [Each bookmark] and it is cut before every page where a top-level bookmark starts; a file holding a whole chapter gets the chapter name. Without bookmarks, click the scissors between the pages where chapters change, or type ranges in the box with semicolons, such as 1-4; 5-9.

What is the difference between a comma and a semicolon in a range?

A comma adds pages to the same file and a semicolon starts a new file. 1-2, 5-6 is one file holding pages 1, 2, 5 and 6, while 1-2; 5-6 is two files of two pages each. 9- means page 9 to the last page.

Why are the split files bigger in total than the original?

Fonts that several pages shared are copied into every split file. The sample PDF in this tool is 204 KB, and splitting it by bookmark into 6 files gives a total of 944 KB. If one file is still large, shrink its photos with Compress PDF.

Do links and bookmarks survive in the split PDFs?

Links to web addresses and annotations such as highlights and notes are kept. Links that jump to another page inside the document, such as contents links, as well as bookmarks and the printed page number setup, are dropped. An item pointing to a left-out page could otherwise drag that page into the result file.

Can I split a password-protected PDF or one with form fields?

No. Password-protected PDFs and PDFs with fill-in fields (forms) or signature fields are not opened. Use a copy with the password removed, or for a form, make a copy with the entries built into the pages using Save as PDF from the print menu, and split that.