Hacks VitaeAll tools
Hacks Vitae Tools
Precision · Instant · Private

Take the pages you need out of a PDF

Take the pages you need out of a PDF and leave the rest behind. You list pages and ranges, and a new file is built from just those, in the order you typed them. It happens in this tab, so a contract or a medical letter never goes anywhere.

SYSTEM ● ONLINE · LOCAL COMPUTE · ZERO UPLOAD
UNIT // PDF.SPLITLIVE
Drop a PDF file
Drag & drop or click to browse
0
Total pages
0
Selected
What it does

It builds a new document from the pages you pick

// In short

Nothing is cut out of your original — that file is left exactly as it was. A fresh, empty PDF is made, the pages you listed are copied into it, and that copy is what downloads. Because the pages are copied rather than re-drawn, text stays text: you can still select it and search it afterwards.

Worth knowing

What travels with a page, and what does not

Bookmarks do not come across. They are not part of a page. The PDF standard keeps them in a document outline, “a tree-structured hierarchy of outline items (sometimes called bookmarks)” held by the document itself (ISO 32000-1:2008, 12.3.3). A new document assembled from copied pages therefore starts without one. Links that pointed to a page you did not take have nothing to point at either.

Removing a page is not redaction. Leaving a page out means the new file does not contain it. It does nothing about words still printed on the pages you kept. Redaction is a separate job, and this tool does not do it.

Check the page numbers you can see. The box counts sheets from 1, which is not always the number printed on the page — a report with a cover and contents often starts its own numbering later.

FAQ

Frequently asked questions

Commas between items, a hyphen for a range: 1, 4, 9-11. Spaces make no difference.
Yes. They are copied in the order you list them, so 3, 1, 2 comes out in that order.
No. It is read, never written to. The download is a new document.
Bookmarks do not: the standard keeps them on the document rather than on a page (ISO 32000-1, 12.3.3), so a new file built from copied pages has none. Links pointing at pages you left out will not work.
Only in the sense that a page you leave out is not in the new file. It does nothing about sensitive text on the pages you keep, and it is not a substitute for proper redaction.
Worth knowing

This extracts pages — it does not burst a file apart

// Answer

One run produces one new PDF containing the pages you listed, in the order you listed them. It does not produce a separate file per page, and it does not touch your original. To get several separate files, run it several times and change the range each time.

That is worth being clear about, because "split" is used for two different jobs. If you wanted forty single-page files out of a forty-page document, this is the slow way to do it.

Syntax

What the range box really accepts, including the traps

Watch the Selected counter as you type. It is the only reliable confirmation that what you typed means what you think. If it says 1 and you expected 12, stop and reread the box.

You typeYou get
1-3, 5, 8-10Pages 1, 2, 3, 5, 8, 9, 10. Spaces are ignored
3, 1, 2Those three pages in that order. Extracting is also reordering
1, 1, 4Page 1 twice, then page 4. Duplicates are allowed on purpose
5-1Nothing. Ranges only count upwards. To reverse, list pages individually: 5, 4, 3, 2, 1
1–3 with a long dashOnly page 1. A word processor may have turned your hyphen into an en dash; only the plain hyphen is read as a range
1 to 3Only page 1. There is no word syntax
12 in a 10-page fileIgnored without complaint. Out-of-range numbers are dropped silently
0 or a negative numberIgnored. Pages are numbered from 1
allNothing. There are no keywords — type the real last page number
1-endOnly page 1. The number before the hyphen is read and the rest ignored
The common need

Removing a page, not keeping one

Often the job is to delete something: a blank final page, a cover sheet, an appendix, a page with the wrong figure on it. The box does not take a delete list. It takes a keep list, so you invert the problem.

  • A ten-page file, drop page 7: type 1-6, 8-10.
  • Drop the first two pages of a forty-page report: type 3-40.
  • Keep only the odd pages of a duplex scan: type them out, 1, 3, 5, 7. There is no every-other-page option.

The Total pages figure at the top of the tool is read from your file, so you always have the real last page number in front of you.

The thing that goes wrong

Why two pages out of a 50 MB file are not 2 MB

Pages do not carry their size on their own. They reference fonts, images and colour data, and those resources come along with the copy. Extract two pages from a long report and the fonts those pages use come along exactly as they were embedded in the original, and are not cut down further here. Extract a page from a scan and you get that page's full-resolution image.

So the new file is smaller, but rarely in proportion to the page count. If the size is the actual problem, follow the extract with compress PDF — and read the limits there first, because a scanned page will not shrink much whatever you do to its structure.

Limits

When not to use this

  • The file is password-protected. Encrypted PDFs are not opened, and you get a short message saying the file could not be read. Remove the password where it was set.
  • The document is digitally signed. The extract is a new file, so the signature does not come with it. Never send an extracted page as proof of a signed original.
  • You need the bookmarks. The outline belongs to the document, not to the pages, so the new file has none.
  • It is a fillable form. Fields may stop working once their pages are copied into a new document. Test the output before sending it to anyone who has to type in it.
  • You need one file per page. Nothing here produces a ZIP. One run, one file.
  • You are trying to redact. Removing a page removes that page, but everything on the pages you keep is still fully there, and text under a box drawn on top of it may still be in the file. Splitting is not redaction.
Worked example

They asked for the signed page, not the contract

  1. Drop the contract. Note the Total pages figure.
  2. Open the file in your normal reader and find the page number you need. The tool counts pages in file order, starting at 1, regardless of printed page numbers or any page labels your reader shows.
  3. Type that number, check Selected reads 1, and extract. The download is named after your original with -pages on the end.
  4. Open the result and look at it before sending. Thirty seconds of checking beats sending the wrong page of a contract.

If you need to put pieces back together afterwards — the signed page plus a new appendix, say — that is merge PDF. If the appendix is currently a set of photos, images to PDF turns them into a document first.

Privacy

Sending less is the point

Extracting pages is usually a privacy act in itself. You are separating the three pages somebody is entitled to see from the thirty they are not: the statement pages without the rest of the year, the medical letter without the full history, the payslip without your colleagues' figures.

It would be a strange act to perform by uploading the whole document to a stranger's server first. Here the file is read by your own browser, the pages are assembled on your own machine, and the result goes straight to your downloads. Nothing is stored, so there is nothing to delete afterwards and no link anybody else could find.

The rest of the set works the same way — see the compression tools or everything at once.

Sources

Where these facts come from

Last reviewed 21 October 2026

4 more

More documents tools

All 7 →

Everything here runs in your browser. Browse all the tools, or start from the home page.

Ask an assistant about this page

Opens in a new tab with this page and the question already filled in.