Split PDF & Extract Pages

This tool pulls selected pages out of a PDF and gives you them as a new document. Type the pages you want — 1-3, 5, 8-10 — and the extract is assembled in your browser and downloaded. The original is never modified and never uploaded, which matters because the documents people split are contracts, statements and scans.

✓ Runs in your browser — nothing uploaded
✓ No sign-up
✓ Free, no limits

Choose a PDFOpened in your browser — the file is never uploaded.

How extracting differs from deleting

A PDF is a container holding page objects and the resources they reference. Extracting is building a new container and copying the pages you asked for into it, in the order you asked for them. Nothing is re-rendered and nothing is recompressed — the pages in the output are the same objects as the pages in the input.

That makes "extract the pages I want" and "delete the pages I don't" the same operation seen from either end, which is why this tool only offers one of them. Keeping pages 1-3 and 7 of a ten-page document is exactly deleting 4-6 and 8-10, and expressing it as what you keep is less error-prone, because you can read the result back before committing to it.

Because pages are copied rather than rebuilt, text stays selectable, vector artwork stays sharp and images keep their resolution. The output is also usually much smaller than the input, since the resources belonging to discarded pages go with them.

Writing the page range

Ranges are one-based and inclusive, separated by commas, and applied in the order you type them — which means this doubles as a reordering tool.

What each spec produces from a 10-page document

1-3, 5        pages 1, 2, 3, 5          the common case
3, 1          pages 3 then 1            order is preserved
5-3           pages 5, 4, 3             a reversed range counts down
2, 2          page 2 twice             duplicates are allowed
1-9           everything but page 10    deleting, expressed as keeping

A reversed range counts backwards rather than being silently corrected to 3-5, because quietly reordering someone's document is a worse failure than doing what they typed. Anything that is not a page number or a range is rejected with an explanation instead of being guessed at, and a page number outside the document is refused rather than clamped.

The summary under the field shows exactly which pages the output will contain before you press anything, written back in the same notation you typed. On a signature page pulled from a forty-page contract, reading that line back is worth the two seconds it costs.

The documents people split

Look at what gets split, and the case for doing it locally makes itself.

It is the signature page of a contract, sent on to someone who does not need the other thirty-nine. It is three months of bank statements pulled out of a year for a mortgage application. It is a single scanned page of a passport or an ID for a visa form. It is one section of a medical file, one invoice out of a batch, one chapter of an unpublished manuscript.

The reason someone is splitting is very often precisely that the rest is confidential — they are extracting the part that can be shared. Uploading the whole document to a server in order to do that inverts the intention completely: to avoid sending thirty-nine pages to a colleague, you have sent all forty to a stranger.

Here the file is read by your browser and the new PDF is written by your browser. Nothing is transmitted. Open the Network tab and split something, or disconnect from Wi-Fi and split anyway — it works either way, because there is nothing at the other end.

Limits worth knowing

Password-protected PDFs are refused rather than silently unlocked. Open the file in any viewer, supply the password, save an unlocked copy, and split that. A tool that stripped the encryption for you would be removing a protection someone applied deliberately.

Bookmarks are not carried over. The clickable outline belongs to the document as a whole rather than to any page, so the extract is a new document without one. Page-level content — annotations, links within the copied pages, and form fields — comes across with the pages themselves.

The original file on your disk is never modified. The extract downloads as a separate file with -pages added to the name, so there is no way to lose the source by mistake.

Everything happens in your device's memory, so a very large scanned document will be slower on an older phone than on a laptop. There is no page or file-size cap beyond that, and no watermark on the result.

One thing worth saying plainly: extracting pages is not redaction. If a page contains something confidential, removing other pages does not remove it, and drawing a black box over text in a PDF editor usually does not either — the text stays in the file underneath the rectangle and can be selected straight out of it. Splitting is the right tool when the sensitive material lives on pages you can drop entirely; it is the wrong tool when it lives inside a page you need to keep.

Frequently asked questions

How do I delete pages from a PDF?

Extract the pages you want to keep. Deleting pages 4-6 from a ten-page document is the same as keeping 1-3 and 7-10, and expressing it as what you keep lets you read the result back before committing. Type 1-3, 7-10 and the summary confirms exactly what the output will contain.

Can I reorder pages while extracting?

Yes — pages come out in the order you type them. 3,1 produces a two-page document with the original page 3 first. A reversed range like 5-3 counts backwards rather than being corrected, because silently reordering a document is worse than doing what was asked.

Is my PDF uploaded to a server?

No. The file is read and the new document is written entirely in your browser, so you can disconnect from the internet after the page loads and it still works. This matters because splitting usually means separating a shareable part from a confidential one.

Does extracting reduce quality?

No. Pages are copied rather than re-rendered, so text stays selectable, vector artwork stays sharp and images keep their original resolution. The extract is usually much smaller than the source, because resources belonging to discarded pages go with them.

Why won't my password-protected PDF open?

Encrypted PDFs are refused rather than silently unlocked. Open the file in a viewer, enter the password, save or print an unlocked copy, and split that instead.

Will bookmarks and form fields survive?

Form fields and annotations attached to the copied pages come across with them. Bookmarks do not — the outline belongs to the document rather than to any page, so the extract is a new document without one.