A 300-page lecture course, an annual report, a thesis: these documents almost always have an interactive table of contents — bookmarks, the “outline” in PDF jargon. When it exists, it does the splitting work for you: each bookmark marks the start of a chapter, and outline extraction produces one file per chapter without you having to hunt down page numbers.
How the tool splits
The principle fits in one sentence: for each bookmark at the chosen level, the output file runs from its page to the page of the next bookmark. Concretely:
- You pick the outline level (1 to 10, level 1 by default — the main chapters; level 2 for subsections).
- The tool reads the bookmark hierarchy and computes the bounds of each chapter.
- You get
document-flare-extracted.zip, with one PDF per chapter, named sequentially (-001,-002…).
What matters in practice
Four implementation details change the outcome — worth knowing upfront:
- Pages before the first bookmark are not extracted. Cover, preface, front matter: everything before the first bookmark at the chosen level lands in no file. If you need it, extract those pages separately with page extraction.
- A bookmark with no target page is ignored. Some bookmarks point to an action or an external link rather than a page: they don’t create a cut.
- Level 2 and beyond: only the first subsection of each parent is read. This is a real limitation of our outline reader: for sub-chapters, stick to level 1, or check the result on a small test before trusting the ZIP.
- No bookmark at the chosen level → explicit error. The tool warns you rather than producing an empty file.
The typical workflow
The most common case: a teacher receives the full textbook PDF and wants to hand out one chapter per week. Level 1, one click, and each ZIP contains a complete chapter with its images and formatting intact — extraction copies pages without recompression. Same logic for an annual report where each department only wants its section, or a legal file where only one exhibit is circulated.
If the document has no bookmarks — common with exports and scans — two alternatives: text-based splitting when a marker word opens each section, or splitting every X pages for a regular cut. The article Bookmarks in a PDF explains how to check and fix an existing table of contents.
Limits
- 1 file, 50 MB, 1,000 pages — the 3,000-page manual will need splitting first.
- Output file names are sequential, not derived from chapter titles: renaming is still needed.
- Processing is local: the document never leaves the browser, which matters for confidential courses or files.
In short: when an outline exists, it’s the fastest and most reliable splitting method — provided you aim at the right level and remember that front-matter pages stay out of the ZIP.
