Split a PDF, or keep just the pages you need

Type the pages you want the way you would in a print dialog. Take them out as one document, or split them into a file each. The PDF is handled in this browser tab and is never uploaded.

How it works

Splitting a PDF is nearly always the last step before sending something. One page of a twelve page statement. The two pages of a contract that were actually signed. A scan where page three came out upside down and has to go. A booklet that needs to become one file per chapter. In every case the pages themselves are fine, and the only thing wrong is which of them are in the file.

This copies the pages you name into a new document, object for object, in the order you named them. Nothing is re-rendered and nothing is re-compressed, so the text is still text, the links still work, and a mixed A4 and Letter document keeps both sizes. That is the difference between splitting and compressing: compressing to a hard size limit eventually has to turn pages into pictures, and this never does.

The page list is the same notation a print dialog uses, because it is the one everybody already knows. Two things it does that a print dialog usually will not: leave the end open, so 8- means page eight to whatever the last page is, and run a span backwards, so 10-1 hands back the document reversed. The order you type is the order you get, and a page written twice is copied twice.

How to use it

1 Choose your PDF.
2 Type the pages you want, like 1-3, 7. The result appears below as you type.
3 Reorder or drop pages there if you want to, or edit the list directly.
4 Take the selection as one PDF, or as one file per page in a ZIP.

Good to know

To delete pages, name the ones you are keeping. 1-4, 6- keeps everything except page five.
The order is yours: 3,1,2 really does put page three first, and 1,1 copies a page twice.
Nothing is re-encoded, so the pages you take out weigh what they weighed. Compress afterwards if there is a size limit.
Bookmarks and fill-in form fields belong to a document rather than to its pages, so they do not survive. What is printed on the page is unchanged.
A password-protected PDF cannot be read. Remove the password in your PDF reader first, then use the copy.

Browser support

Chrome 94 and later
Firefox 93 and later
Safari 16.4 and later
Edge 94 and later

These are the oldest versions with everything the tool uses. They are worked out from the tool itself, so they change when it does.

Version numbers checked on 2026-09-04.

A PDF is a bag of objects, which is why taking pages out costs nothing

Most file formats are a stream you read from the start. A PDF is not. It is a collection of numbered objects, and at the very end sits a cross-reference table saying which byte offset each object lives at. A page is one of those objects. It holds references to the fonts it uses, the images it draws, the text operators that place its glyphs, and a box giving its size. Nothing about a page says which document it belongs to or which page number it is, because that lives in a separate tree of nodes whose leaves point back at the pages.

That structure is the reason splitting can be free. Extracting page seven means copying that object and everything it refers to into a new document, then building a one-leaf page tree that points at it. The glyph positions are untouched. The images are the same compressed streams they already were, at the same resolution. Nothing is decoded and nothing is re-encoded, so there is nothing to lose. Compare that with compressing to a hard size target, which eventually has no choice but to rasterise a page into a picture of itself and throw the text away.

It also means the page size travels with the page rather than with the document. A file where page one is A4 and page two is US Letter is completely ordinary, and each page carries its own MediaBox to say so. A splitter that normalised everything to one size, or that assumed the first page spoke for the rest, would silently reflow scans that were fine. Copying the object keeps the box, so a mixed document stays mixed.

Two things genuinely do not survive, and it is worth knowing why rather than being surprised later. Bookmarks are a separate outline tree with its own destinations, pointing at pages by reference; keep four pages out of forty and most of those destinations now point at nothing, so the outline is dropped rather than left broken. Form fields are the same story: the fields live in a document-wide AcroForm structure while only their widgets sit on the page, so an extracted page still shows the text that was typed into it, drawn exactly where it was, but there is no longer a field behind it to edit. The page looks identical and is no longer fillable.

The last piece is the page list itself. It uses print-dialog notation because that is a notation people already have in their heads, and it turns out to express more than a print dialog uses it for. Order is meaningful, so 3,1,2 is a reordering rather than a selection. Repetition is meaningful, so 1,1 copies a page twice. Direction is meaningful, so 10-1 walks the range backwards and hands back a reversed document. One grammar covers extracting, deleting, reordering and reversing, which is why this tool has one text box instead of four modes.

Frequently asked questions

How do I split a PDF without uploading it?

Choose the file here. It is parsed inside your browser tab by your own device, and the pages you asked for are handed straight back as a download. You can confirm it in your browser’s developer tools, on the Network tab: no request carries the document.

How do I delete pages from a PDF?

Name the pages you want to keep rather than the ones you want gone. For a ten page file without page five, that is 1-4, 6-. The result is a new document containing only what you named.

Can I change the order of the pages?

Yes. The order you type is the order you get, so 3,1,2 puts page three first. The arrows next to the result do the same thing by rewriting the list for you, and a backwards span like 10-1 reverses the whole document at once.

Will the text still be selectable?

Yes. Pages are copied rather than re-rendered, so text stays text and can still be searched, selected and read aloud by a screen reader. Images keep their original resolution too.

Can I get one file per page?

Yes. The second button writes a single-page PDF for every page in your selection and hands them over as one ZIP, named after the original with the page number appended.

Can I split a password-protected PDF?

No. An encrypted PDF cannot be read without its password. Open it in your PDF reader, save a copy without the password, and use that copy here.

Your files are processed on your device and are not uploaded. You can verify this in your browser’s developer tools under the Network tab.