What this tool does
This is the mirror image of splitting and it deserves its own tool, because the two are psychologically opposite. Splitting can fail loudly with nothing lost. Deleting cannot, so the interface is built the other way round: the destructive action is never the default, the exact page numbers are stated before you commit, and the opposite operation is always one click away.
How it works
Deleting pages is a set operation, and the set you type is the set that goes. Every other page is kept, which is worth being explicit about, because the natural reading of remove pages 3 and 7 is sometimes the opposite: some people mean keep only 3 and 7. The tool offers both, labelled as what they do, so the intent is chosen rather than inferred from ambiguous wording.
Because there is no undo, the confirmation line is the interface. It reads back the pages that will be removed and the pages that will survive, both as numbers, before anything happens. The output is a new file and the original on your computer is never touched, so even a badly-chosen range is recoverable by starting again, but reporting it in advance is still the difference between a tool you trust and one you check afterwards.
The keep-instead-of-delete action is the safety net. It is the same parser and the same page list, run in the opposite direction, and it is offered on the same control rather than buried in a separate tool. This exists because the destructive direction has no reversal in the file itself: once the pages are gone from a saved copy, they are gone.
Blank pages are the usual reason people look for this. A scanner run on a duplex tray adds a blank verso for every odd page, and a fax or a print-to-PDF of a two-sided document does the same. Auto-detect finds pages with no content or content that draws nothing visible, but it is a heuristic and the tool says where it can go wrong: a page that is genuinely blank apart from a white background image will read as having content, and a page with a coloured header will read as having content when it looks blank to you.
Worked example
A duplex scan produced a 40-page PDF where every odd page has a blank reverse, and you need the 20 real pages as one document.
- Try auto-detect, which finds the 20 pages that draw nothing
- Check the preview: it lists the pages it will remove, which should be exactly the 20 even numbers
- If auto-detect is wrong on a page, add or remove that page number by hand from the expression
- Run the removal. The output is a new 20-page file and the original is untouched
A clean 20-page PDF with no interleaved blanks, produced without uploading a document that may well have been somebody's payslip.
Accuracy and limitations
- The original file on your computer is never modified, so there is nothing to undo and no data is lost from the source. Getting the range wrong means running it again, not recovering a file.
- Auto-detecting blank pages is a heuristic. A page whose only content is a white image reads as non-blank, and a page with a header or a page number on it reads as non-blank even if it looks empty.
- Deleting a page can break the internal links and bookmarks that pointed at it, which is normal for any tool that removes pages.
- A password-protected PDF cannot be opened here.
Frequently asked questions
- How do I remove a page from a PDF?
- Enter the page numbers you want gone, using commas and dashes, such as 3,7,11-14. The line under the input reads back both the pages that will be removed and the ones that will be kept, before you run anything, so the range can be confirmed rather than assumed.
- Can I keep only certain pages and remove everything else?
- Yes, and that is the other mode on the same control. Enter the pages you want to keep, such as 1,4,9, and everything else is removed. It is offered next to the delete mode rather than in a separate tool, because the two are the same operation in opposite directions and people describe either one as deleting pages.
- How do I remove blank pages from a scanned PDF?
- Use auto-detect, which looks for pages that draw nothing visible and offers to remove them. It is a heuristic rather than a certainty, so check the preview: a page whose only content is a white image will read as non-blank, and a page carrying a header or a page number will read as non-blank even if it looks empty to you. Add or remove individual page numbers by hand to correct it.
- Is deleting pages reversible?
- The output is a new file and the original on your computer is never modified, so you can always start again from the source. Within a saved copy, though, the removed pages are genuinely gone, which is why the tool shows the exact page numbers before it acts and offers the opposite operation next to the destructive one.
- Will removing pages break the document's links or bookmarks?
- It can. Anything that pointed at a removed page, including bookmarks and internal hyperlinks, has nothing left to point at. This is normal for every tool that removes pages rather than a fault with this one, and it is one of the reasons to keep a copy of the original.
- Are my PDFs uploaded when I remove pages?
- No. The document is opened, rebuilt without those pages and handed back as a download, all inside your browser. The file itself is never sent to anyone, which is worth having for the documents people most often clean up, such as bank statements and payslips.