Remove pages from a PDF without uploading it

The reason people trim a PDF is usually the reason they should not upload it. A twelve page contract where only the schedule is relevant. A bank statement where one transaction is the evidence and the rest is nobody's business. A report with an internal appendix that was never meant to go outside.

In every one of those cases, the pages being removed are the sensitive ones. Uploading the whole document to strip them means sending the confidential part to a third party in order to delete it, which is precisely backwards. Doing it locally is the only version of this task that makes sense.

The thumbnail grid matters more than it sounds, because trimming by page number is where mistakes happen. Somebody says remove pages four to six, the document has an unnumbered cover, and the pages that go are five to seven. Seeing what you are removing removes an entire category of error, and it is the sort of error nobody catches afterwards because the document looks perfectly reasonable.

There is one important thing this cannot do, and it is worth being blunt about. Removing a page removes the page. It does not redact: text hidden under a black rectangle is still text, and a name covered by a shape is still in the file for anyone who selects it. Proper redaction removes the underlying content, and a surprising number of published documents have leaked precisely because somebody drew a box over something.

So the safe pattern for anything sensitive is to remove whole pages rather than to obscure parts of them, and where only part of a page can go, to rebuild that page rather than cover it. Less elegant, considerably harder to get wrong.

Everything runs in this page. No account, no upload, no retention policy to read, because there is nothing to retain.