PDF splitter for pages, ranges and bookmarks

One PDF per page for scanning workflows; custom ranges when you know the chapters ("1-12, 13-40"); every N pages for splitting a long report into equal parts; a single extract when you want one slice and nothing else; and by bookmarks when the document already carries its own structure.

At every TOP-LEVEL entry in the outline — chapter headings, not the sub-sections nested under them — and each output file is named after the entry it starts with. A document with twelve chapters and eighty sub-headings gives twelve files, not ninety-two.

One output means one PDF, straight to your downloads. Several outputs are packed into a ZIP, because a browser only reliably hands over one file per click. The pieces arrive with the names your pattern produced, and you can see those names under the pattern field before you click. If the pattern would give several pieces the same name, each one gets its piece number — otherwise they would overwrite each other in the archive.

Yes, through the filename pattern. The tokens are {name} for the source file, {idx} for the piece number, {page}, {start} and {end} for page numbers, and {title} for the bookmark a piece starts with. "{name}-{start}-{end}" gives report-13-40.pdf; "{idx}. {title}" gives 3. Methods.pdf.

Each output keeps the part of the outline that belongs to the pages it contains, so a reader still gets a working table of contents inside the piece. Page labels travel too: a preface split off on its own still numbers i, ii, iii rather than restarting at 1.