How to split a PDF and extract the pages you need

Sometimes you have a 40-page download and you need pages 4 through 6. Sometimes you have a stack of scanned receipts in one PDF and you want each receipt as its own file. Both are splitting, and they want different outputs. A range should come back as one PDF that still reads in order. A stack of receipts should come back as many PDFs, usually in a zip so the browser is not asked to save forty files one by one.

PDF Split does those two jobs and not a third one. It does not rearrange pages into an arbitrary order inside a single click, and it does not split by chapter heading, because most PDFs do not store a reliable heading structure the page can trust. You name pages by number, or you take every page.

The file is read locally with pdf-lib. As soon as you add a PDF, the page counts it and shows the number. That count is the number of pages in the file, not whatever a footer claims. If a footer says “Page 3 of 10” and the file has 8 pages, trust the 8.

Range mode starts with the full span filled in, such as 1-8, so a successful read shows you a legal value immediately. Edit it down. Page numbers start at 1. Commas separate parts. A hyphen marks a span.

1-3,5 means pages 1, 2, 3, and 5. The result is one PDF containing those pages. The selected pages are made unique and then sorted from low to high. Writing 5,1-2 does not put page 5 first. It produces pages 1, 2, and 5. If you need page 5 in front, extract page 5 and pages 1-2 as two operations, then put them in order with PDF Merge.

A token that is not a number or a span, such as a letter, is rejected with the token quoted in the error. A number larger than the page count is rejected and the message tells you how many pages the file has. A backwards span such as 5-2 is rejected and the message tells you to write 2-5. Blank input asks you to enter a range.

The download name is the original file name with the extension replaced by extracted.pdf. A file called packet.pdf becomes packet.extracted.pdf. Your original is not overwritten.

Single-page mode ignores the range box. Each page is written to its own PDF. The names look like packet-page-01.pdf, with the number padded so the files sort in order. A file of a single page downloads that one PDF directly. A file of several pages is packed with JSZip in the browser and downloaded as packet.pages.zip. Unzip it to get the individual PDFs. A reader that does not open zip files is not failing to read the PDF; it is looking at the archive.

This mode is the one to use for a scan that should have been separate files. It is a blunt tool for a chapter extract. You would get every page, including the ones you meant to drop.

Drop one PDF. The control holds a single file; a new drop replaces the previous one. Wait for the page count so you know the read succeeded. Choose the mode. Narrow the range if you are extracting. Click Split and download.

If the count never appears and you see an error instead, the file did not open. Password-protected PDFs fail here. Remove the password in a reader that already knows it, then split the unlocked copy. QuickConvert does not take the password. A damaged file fails the same way, with a message that it could not be read. A file that is not a PDF is rejected by the drop zone before that.

The 80 MB limit is there because the whole file sits in memory. A long color scan can be under that limit and still make a phone struggle, especially in single-page mode, which builds a PDF per page and then a zip. On a low-memory device, extract the range you need instead of exploding every page.

Rename packet.extracted.pdf to something that says what the extract is. “Pages 4-6” in the name will outlive your memory of the task. If you extracted several ranges as separate files and now need them as one packet in a specific order, merge those extracts and use the arrows to set the order. Do not expect the split tool to remember a custom sequence.

If the reason you were splitting was file size rather than page selection, check whether the pages are photos. PDF Compress can shrink embedded images without removing pages. Splitting a text PDF into single pages rarely makes the set easier to send, and the zip of many small text PDFs can be a similar size to the original.

The bytes stay on the device. Counting pages, parsing the range, copying pages, and building the zip do not post the document to QuickConvert. You can read more about what the surrounding page still does — analytics, ads, hosting logs — in privacy with online file tools. Open the download before you delete anything. A split that you have not looked at is how a missing page gets discovered by someone else.