You have a 400-page manual, a thesis or an e-book, and you need each chapter as its own file. If the PDF has a bookmarks panel, the document already knows where every chapter starts. Split by bookmarks reads that list and cuts the file at each entry, so you don't have to look up a single page number.
It runs in your browser. The file isn't uploaded.
What a PDF bookmark actually is
The sidebar you see in Adobe Acrobat Reader or Preview on a Mac is called the outline in the PDF specification. It's a tree of entries stored in the file, separate from the pages. Each entry has a title ("Chapter 3: Installation") and a destination, which points to a page and usually a position on it. Entries can have children, so a chapter can hold sections and a section can hold subsections.
The title and the page are linked only by that pointer. A bookmark doesn't contain its chapter. It says "start reading here", nothing more. That's why splitting by bookmarks works at page level: a file starts on the page a bookmark points to and runs until the page before the next one.
Some destinations aren't page pointers at all but names, like "chap3", which are looked up in a separate table elsewhere in the file. Documents made with LaTeX do this a lot. The tool resolves those names too, so they split like any other bookmark.
Why some PDFs have no bookmarks
The outline only exists if the software that made the PDF wrote one. Plenty of software doesn't.
- Microsoft Word on Windows adds them when you save as PDF with the option to create bookmarks from headings turned on (File, Save As, PDF, Options). If your headings use Word's built-in Heading styles, you get a proper tree.
- Printing to PDF usually drops them, because printing only sends what's drawn on the page.
- Scans have none. A scanner produces pictures of pages, not structure.
- Some tools throw the outline away when they merge or edit a file.
If the tool shows "No bookmarks found", one of those happened. You have a few ways out. Split by text can split wherever a heading such as "Chapter" appears on a page. Create bookmarks lets you add an outline yourself and then come back here. And for plain page numbers, Split PDF does the job (the guide to splitting a PDF goes through its modes).
There's a rarer message too: "This PDF has bookmarks, but none of them point to a page in this file." You'll see it when a document was cut out of a bigger one and its bookmarks still point at pages that are no longer there.
Splitting a PDF by bookmarks step by step
- Open Split by bookmarks and drop your PDF on the upload box. A password-protected file asks for its password first.
- Wait for "Reading bookmarks...". You then get one row per output file, with its title, a thumbnail of its first page and its page range.
- Set "Bookmark level". "1 (top level)" splits at chapters only.
- Untick any rows you don't want. The "Select none" and "Select all" button helps with long lists.
- Click the button in the header, which tells you how many files you'll get, for example "Split into 12 file(s)".
One file downloads as a PDF. Several come together as a ZIP named after the original, ending in -by-bookmarks.zip.
Choosing the right bookmark level
Level 1 is the top of the tree, usually chapters or parts. Level 2 adds the entries one step down, and so on, up to the deepest level the file has. Deeper levels don't replace higher ones. At level 2 you split at every chapter and at every section.
That has a side effect. A chapter heading and its first section often point to the same page. Two bookmarks on one page would produce a file with no pages, so the second one is folded into the first, and a note above the list tells you how many were combined. Nothing goes missing; the file takes the name of the first bookmark.
The "Contains text (optional)" box narrows things down. Type Chapter and only bookmarks with "chapter" in the title start a new file (upper and lower case are treated the same). This helps with textbooks where the top level mixes chapters with a preface, an index and the answers to exercises. Bookmarks that don't match don't cause a split, so their pages stay in the file before them.
I'd start at level 1 with no filter and read the list. Go deeper only if a chapter is still too big for what you need it for.
The options worth knowing about
"More options" holds four more settings.
"Include extra page at the end" adds the page where the next bookmark starts to the end of each file. Bookmarks point to pages, not paragraphs. If chapter 4 ends halfway down page 88 and chapter 5 starts lower on that same page, chapter 4's file would otherwise stop at page 87 and lose its last few paragraphs. The cost is that page 88 appears in both files.
"Keep pages before the first bookmark" is ticked by default. Covers, title pages and contents pages often have no bookmark of their own, and this saves them as a separate file called "Pages before the first bookmark" instead of dropping them. Untick it if you only want the chapters.
"Discard bookmarks" removes the outline from the new files. Left unticked, each file keeps the bookmarks that point to its own pages, so a chapter you split off still has its section bookmarks.
"Output file names" takes a pattern. Without one, each file is named after its bookmark with a number in front, like 03-Installation.pdf, so the files sort in the right order in a folder. A pattern can use [FILENUMBER], [BOOKMARK_NAME], [BASENAME], [CURRENTPAGE] and [TIMESTAMP]. Characters that Windows and macOS refuse in file names are cleaned out of bookmark titles automatically.
What it won't do
It splits at whole pages, never in the middle of one. It can't split a scan that has no bookmarks, and it doesn't read the headings printed on the page (that's what Split by text is for). It also trusts the bookmarks completely. If they were left over from an older version of the document and point one page too early, every file will start one page too early.
Each new file carries its own copy of the fonts and images its pages use, so the parts together are often a little bigger than the original. That's normal.
When they are off, the quickest fix is usually Split PDF with "Pick split points", using the bookmark list in your reader as a rough guide to where each chapter really starts.
Also useful: Split by text · Create bookmarks · Split PDF
