Delete Pages from a PDF
Click the thumbnails you do not want and they are skipped on export. Marked pages get a red badge; mark one by mistake and click again to undo.
Click a thumbnail to mark it for removal, click again to undo. Marked pages are skipped on export.
How to delete pages from a PDF
- Drop the PDF into the dashed box, or click it to pick a file. It is parsed into thumbnails locally, right away.
- Click a page to mark it for removal — a red badge appears in its corner. Click again to unmark. On long documents it is faster to select all and then unmark the pages you want to keep.
- Press Delete pages and export. The new file downloads automatically; the original is untouched.
Will the file get smaller? Almost never
This is the most common misunderstanding about page deletion, so it is worth being blunt: removing pages usually does not reduce file size. A PDF is an object tree. A page is one node on that tree, while the images, fonts and content streams it draws on are separate objects stored elsewhere in the file.
Deleting a page detaches the node. The objects it referenced stay in the file, unreferenced but present. Since images dominate PDF size, deleting ten text pages might save a few kilobytes, and deleting a page with a 5 MB scan may save nothing at all, because that image's object is still sitting there. Making a file smaller is the job of compression, which resamples images — a completely different operation with a real quality trade-off.
In one line: deleting pages changes what the document contains, not how big it is. People conflate the two constantly, and then conclude the tool is broken when the size does not move.
Can I get the pages back?
Yes, as long as you have not closed the tab. Reset clears every mark and restores the original order. And even after exporting, the file on your disk is still complete — this tool never edits the source, it only produces a new file.
When this is the right tool
A scanned contract with a blank page and a cover sheet mixed in. The back sides that a double-sided scan added. The old version of a page in a CV attachment. The acknowledgements section that has to come out before a paper goes to review. The pattern is always the same: most of the document stays, a handful of pages go. That is exactly when deleting beats rebuilding.
Delete Pages or Extract Pages?
The two are complementary and the rule is simple: few pages to remove — delete; few pages to keep — extract. Removing three pages out of a hundred is less work than selecting ninety-seven. Both produce structurally identical output: the surviving pages are carried over untouched, at full quality.
What it does technically
The page reads the document with pdf-lib in the browser and performs the delete at the page tree
level: the removed pages are taken out of the /Kids array of the /Pages
node and /Count is corrected to match. Surviving pages are never decoded or
redrawn, which is why quality and byte layout are preserved.
That also explains the two hard limits. An encrypted document's page tree cannot be read, so it cannot be processed. A damaged document has an incomplete page tree, so parsing fails outright. In both cases you get the reason in plain language instead of a spinner followed by an unreadable error.