Merge PDF
Combine PDFs and reorder pages by dragging. Nothing is uploaded.
Extract page ranges or break a document into single pages.
The range field accepts the same syntax as a print dialog, so it should already be familiar.
| You type | You get |
|---|---|
| 1-5 | Pages 1 through 5 as one document |
| 1-5, 12, 20-24 | Three separate documents |
| 1-5; 12; 20-24 | The same three, semicolons also work |
| 3- | Page 3 through the end |
| -7 | The beginning through page 7 |
| 1,3,5,7 | Four single-page documents |
| odd | Every odd-numbered page |
| even | Every even-numbered page |
Odd and even are more useful than they look. A duplex scan produced on a single-sided scanner gives you all the fronts in one file and all the backs — usually reversed — in another. Extract, reverse, and interleave with the Merge tool to reconstruct the document.
Quality, no. Pages are copied as complete objects with their fonts, images and color profiles intact. There is no re-encoding step, so an extracted page is identical to the same page in the original.
File size is less intuitive. Extracting 10 pages from a 100-page document does not produce a file one tenth the size. Shared resources — an embedded font used throughout, a logo on every page — must be copied into each output file. Split a 50 MB document into 100 single pages and the total can exceed the original.
This tool subsets embedded fonts to the glyphs each output actually uses, which recovers most of that overhead. Run the results through Compress PDF if size still matters.
Extract everything except the pages you want gone. To remove page 7 from a 20-page document, request 1-6, 8-20 as a single range set, and the result is a 19-page document.
The tool has a dedicated "remove pages" mode that inverts the selection for you, so you can simply type the pages to delete. It is the same operation stated the other way around, and it is the more natural way to think about redacting an internal cover sheet or dropping a blank scan page.
Yes. A scanned PDF is a container holding one image per page, and splitting it moves those images into new containers without touching the pixels.
What splitting will not do is find the boundaries for you. If you scanned twenty invoices into one 60-page file, the tool cannot tell where invoice 4 ends and invoice 5 begins — that requires reading the content. Render the pages to images with PDF to JPG, or scroll the thumbnails, to identify the boundaries, then enter the ranges.
Yes. Type the page number on its own, or use the per-page thumbnail download button.
No. The file on your disk is opened read-only and never modified. Every result is a new document.
No. Documents of thousands of pages split without trouble, limited only by device memory.
Links pointing to pages inside the extracted range are preserved. Links to pages that did not make it into the output are dropped, since their targets no longer exist.
Only with the password. Encryption is applied to the document structure, so it must be decrypted before any page can be read.
Use the "split every N pages" mode with N set to half the page count, or type two ranges — 1-25 and 26-50 for a 50-page document.
No. Parsing and writing both happen in a Web Worker in this tab.