Split vs extract: getting exactly the pages you need

Extract and split are two different jobs sharing one word, which is why people end up with the wrong file. Extracting means pulling selected pages out of a document into one new PDF. Splitting means breaking a document into several files. Split PDF does both, and you steer between them with two independent controls: the page ranges decide which pages you get, and the output mode decides how many files they arrive in.

The rule that predicts your output

Count the comma-separated groups in the pages field, then read the output mode. Tapping pages in the preview grid rewrites that field for you, and consecutive pages collapse into one hyphenated range, so trust what the field says rather than how many pages you tapped.

One PDF returns a single document, however many groups you entered. That is extraction.

Separate files returns one PDF for each group, packaged in a ZIP. That is splitting.

The part that trips people up is what counts as a group. 2, 5, 9 is three groups. 2-6 is one group that happens to contain five pages, so Separate files returns a ZIP with a single five-page PDF inside it. The full range syntax, including open-ended ranges such as 8-, is in the main Split PDF guide.

Three situations, three answers

A 40-page agreement where only the signature pages need to go out. This is extraction. Type 38-40, choose One PDF, and send the three-page result. The recipient gets one attachment holding only what concerns them, and the ranges never needed to be split apart.

Twelve pages of scanned invoices, two pages per invoice, that a portal wants uploaded one invoice at a time. This is splitting. Type 1-2, 3-4, 5-6, 7-8, 9-10, 11-12, choose Separate files, and the ZIP holds six PDFs, one per invoice. Choosing One PDF here gives you back a document almost identical to the one you started with.

A report going to three different reviewers. This is splitting into deliberate groups rather than into single pages. Ranges of 1-2, 3-18, 19- with Separate files return the cover, the body, and the appendix as three files you can send separately. The same ranges with One PDF would rebuild nearly the whole report, which is rarely the point.

How to get either one

Why the wrong choice looks like a bug

Two reports arrive regularly, and neither is a failure. The first is a ZIP that contains one PDF: that happens when Separate files meets a single continuous range, because there was only ever one group to separate. The second is one document when six were expected: that is One PDF quietly doing exactly what it promises.

Both are fixed with Split again, which keeps the file loaded: change the output mode, run the job once more, and there is no second upload. It is worth checking the mode before you download rather than after you have sent the file on.

One more source of surprise is page numbering. The tool works from the page numbers in the preview, not from numbers printed inside the document. A report whose first content page is labelled page 1 may sit at page 3 in the PDF, and a range typed from the printed numbers will be off by that offset.

Extract, delete, reorder: three different tools

Extraction picks the pages you want to keep. If your real goal is a copy of the document with two pages gone, listing every page you want to keep is the long way around, and Delete PDF Pages does it directly. If the pages are correct but in the wrong order, Organize PDF does that directly: drag the pages into place and download the rebuilt document. And if you split a document, worked on the pieces, and now want it whole again, the merging guide covers putting them back in a set order.

Splitting never rewrites what is on a page. It copies whole pages into new files, so a sideways page stays sideways and a scan stays a scan.

Before you upload either way

Both modes run on our server, so the PDF travels over HTTPS and is used only for that job. Uploads and results are deleted automatically about an hour later, which is why you should download before walking away. There is no account, no watermark, and no daily limit.

Automatic deletion does not override a workplace, school, legal, or healthcare rule that forbids third-party processing. When the rule is local only, use approved offline software. The guide to what happens to an uploaded PDF covers the questions worth asking before any server-side tool touches a document.

When this won’t work

• The PDF is over 50 MB. Split takes one input file and the upload cap applies to it, whichever output mode you pick. Use trusted desktop software for a larger document.
• The file needs a password to open. Make an unlocked working copy with Unlock PDF first, using the password you already know. A forgotten password cannot be recovered by us or by anyone selling a recovery service for a properly encrypted file.
• You need pages reordered rather than separated. One PDF copies the groups in the order you type them, so 3, 1, 2 does come back in that order. For anything longer than a few pages, Organize PDF does the same job by dragging.
• You want to keep everything except a few pages. Extraction works, but deleting the unwanted pages is fewer steps and less error-prone on a long document.
• The document cannot leave your device. Deletion after an hour is still an upload. Use approved local software when policy says so.