How to Split a PDF: Ranges, Every N Pages, or Pick-and-Choose

Split PDF gives you three different ways to pull pages out of a document, because "split a PDF" means different things depending on what you're actually trying to do. Here's each mode, what it produces, and the two failure cases worth knowing about before you rely on it.

Step 1 - drop the PDF

One document at a time, up to 80MB, the same ceiling merge PDF uses, since a split holds the source document plus every output document in memory at once. Every page renders as a thumbnail; on a genuinely long document, over 50 pages, a progress line appears while thumbnails render, since that's the point rendering visibly takes a moment rather than finishing instantly. "Custom ranges" is the default mode, with the input pre-filled to cover every page in the document, so the field always starts with something sensible rather than empty.

Mode 1 - custom ranges

Type something like 1-3, 5, 8-10, comma-separated, and each range becomes its own output PDF: on a document with, say, 12 pages, that exact string produces three files, one with pages 1 through 3, one with just page 5, one with pages 8 through 10. This is the mode to reach for when you know exactly which page numbers you want, and want more than one output file out of a single pass. Ranges can be typed in any order, they don't need to be sequential, and a single page number on its own, without a hyphen, is read as a one-page range.

Mode 2 - every N pages

Set a number and every chunk of that size becomes its own file: N=1 turns every single page into its own PDF, N=10 splits a document into consecutive 10-page files. On a 45-page document with N set to 10, that produces five files, four full 10-page files and one final 5-page file holding whatever's left over, nothing dropped, nothing padded to force an even split. Useful for breaking a long scanned document into roughly equal pieces without typing out every range by hand.

Mode 3 - pick pages

Click individual thumbnails to select or deselect them, keyboard-operable as well as clickable, and Split produces one PDF containing exactly the pages you picked, in their original order regardless of the order you clicked them in. This is the mode for "I need these four specific pages, not a range."

Step 4 - click Split and download

One output downloads directly as a PDF. More than one downloads together as a zip, and each file inside is named by the pages it holds, such as pages-5-8.pdf, and the result list labels it "Pages 5-8 (4 pages)", so you're not left guessing which download is which. On the success screen, each output is also listed individually with its own download button, in case you only want one of several files right away rather than the whole zip.

Rendering a genuinely long document

A document over 50 pages shows a visible "Rendering page N of total" line while thumbnails build, rather than a plain spinner that gives no sense of progress on something that's going to take a few extra seconds. Once thumbnails finish, the rest of the flow, picking a mode, setting parameters, clicking Split, works exactly the same regardless of whether the source document was 5 pages or 500.

When it can't proceed

A password-protected PDF is detected up front and blocks the whole operation with a plain message rather than a partial or garbled result, since splitting needs to read the document's real structure first. A corrupt or unreadable file gets the same honest treatment. Unlike merging, which can skip one bad file in a batch and keep going with the rest, splitting only ever takes one document, so an unreadable file blocks the whole operation rather than producing a partial result. An invalid range string, page numbers out of order or beyond the document's actual length, is rejected as a whole with a message telling you to check the numbers, rather than silently splitting on whatever part of it happened to parse.

The file stays loaded if you need to fix something

If a range string comes back invalid, the document and its thumbnails stay right where they were, you're not sent back to re-upload the whole file just to fix a typo in the page numbers. Correct the range and click Split again; the file only needs to be dropped in once per session.

A real tradeoff worth knowing about

Splitting copies pages exactly as they are, no re-rendering, no quality loss. But each output file needs its own copy of anything that page depends on, an embedded font, a repeated header image, so splitting into many small files, especially Every N pages with a small N, can leave the combined total size larger than the original document, since shared resources get duplicated into every output rather than shared once. The tool shows a plain notice when this applies, it isn't a bug, it's how independent PDF files necessarily work.

Which mode to reach for

If you already know the exact page numbers you want, ranges is the fastest path, type them once and get every output in one pass. If you want the document broken into roughly equal, arbitrary-sized pieces without looking up specific page numbers, every N pages does that automatically. If what you need doesn't map to a clean number pattern at all, a handful of specific pages scattered through a long document, pick pages by clicking directly on the thumbnails that matter and ignoring the rest.

Try it on a real PDF, pick the mode that matches what you're after, and see the exact pages you get before deciding it's right.