Extracting pages from a PDF is not the same thing as splitting one. When you extract pages from a PDF here, the source file never changes. You choose the pages you want, type "1, 4, 9-12" or tap them on screen, and the tool builds a brand new file containing only those pages, in that order. The original stays exactly as it was, still on your device, still whole. Nothing about this tool touches it. That distinction matters more than it sounds: some tools quietly modify or discard the source, and this one never does.
How to extract pages from a PDF
- Upload the PDF you want to pull pages from.
- Tap the pages you want on the thumbnail grid, or type a range like "1, 4, 9-12" into the page field above it.
- Check the count above the grid, it confirms exactly how many pages are selected before you extract anything.
- Choose whether you want one combined file or a separate file per selected page.
- Tap "Extract" and download. Your original file stays untouched on your device the whole time.
When you need this
Pulling one invoice out of a monthly statement is a routine case: you need just page 14 of a 30-page bank statement for an expense report, and the other 29 pages are irrelevant to that claim. Extracting just the signature page of a contract, the final page after everyone has signed, is another regular one, useful when you need to send or file proof of signature without circulating the entire agreement again. It also comes up when pulling the results or summary page out of a long lab or audit report to share on its own, or lifting a single chapter or article out of a scanned book or journal PDF without extracting the whole volume. Job applicants use it to pull one reference letter out of a folder of several without sending the others, and property managers use it to extract just the lease clauses relevant to a specific dispute instead of forwarding the entire signed agreement.
The detail that matters
The page field accepts a comma-separated list, and each entry is either a single page number or a range written with a hyphen. Typing "1, 4, 9-12" selects six pages total: page 1, page 4, and pages 9 through 12. Order in the field does not matter and the output always keeps pages in their original document order regardless of how you typed them. A reversed range like "12-9" is corrected automatically rather than rejected, a page number beyond your document's actual page count is silently ignored instead of throwing an error, and overlapping entries like "1-5, 3-7" simply collapse into pages 1 through 7 once each, never duplicated in the result.
A worked example: a 45-page annual report has its cover page on page 1, the financial summary on pages 9 through 12, and the auditor's signature on page 45. Typing "1, 9-12, 45" extracts exactly those six pages into one new six-page file in the original order, cover first, then the four summary pages, then the signature last, regardless of the order the numbers were typed in.
What survives extraction and what does not comes down to where each thing actually lives inside a PDF. Annotations, highlights, comments, drawn shapes, and links are stored as part of the page itself, so they copy over automatically with the page they belong to. Rotation works the same way: a page you already rotated keeps that rotation in the extracted file. Bookmarks are different. The document outline that powers a table of contents is a separate, top-level structure that is not attached to any one page, so a new file built from a handful of extracted pages never carries that navigation over, even though the pages themselves are complete.
Fillable forms sit in between. A PDF form is not just the visible boxes you see and click into, it is also a document-level registry, the AcroForm, that connects every field by name so the whole thing behaves as one form. Extracting a page keeps the visible box, since that part lives on the page, but does not rebuild the AcroForm registry behind it. A form field can look completely normal in an extracted page and still not fill in reliably, so treat any form page you extract as something to verify before you rely on it.
None of this is unique to extraction, the same rules apply to splitting and merging on this site, since all three build a new document from copied pages rather than editing the original in place. What makes it worth spelling out here specifically is that extraction is usually the tool people reach for when only one or two pages matter, a signature page, a single form, a specific clause, exactly the cases where a form field silently not working is most likely to actually cause a problem later.
How this compares to typical online extract tools
| Feature | This tool | Typical online compressor |
|---|---|---|
| Explains what carries over: annotations, rotation, bookmarks, forms | Rarely disclosed at all | |
| Guarantees the original file is never modified | Usually, but rarely stated outright | |
| File uploaded to a server | ||
| Watermark added | Sometimes, on free tiers | |
| Signup required | Often, past a certain file size |
Limits
Extraction copies existing page data instead of decoding and re-encoding it, so it is fast regardless of how many pages you select. It needs no Web Worker or special browser feature, so it runs on effectively any current browser. Only one source file can be open at a time, and the practical ceiling is your device's memory while that file is loaded: sources above roughly 35MB can slow down or struggle on an older phone's browser tab, while a laptop or desktop handles considerably larger files without trouble. There is no cap on how many individual pages you can select beyond the document's own page count, so selecting 200 out of a 400-page file works the same way as selecting two out of ten.
Frequently asked questions
Can I select pages that are not next to each other, like 2 and 15?
Yes, tap each one individually, or type them into the range field separated by commas: "2, 15" selects exactly those two pages and skips everything in between.
What exactly does typing 9-12 select?
All four pages from 9 through 12 inclusive: 9, 10, 11, and 12. Combine it with other entries using commas, like "1, 4, 9-12", to build up any combination.
Do highlights or comments I added to a page survive extraction?
Yes, page-level annotations are stored as part of that page's own data, so they travel with it into the extracted file automatically.
Will a fillable form field still work in the extracted page?
Not reliably. The visible box usually survives, but the underlying structure that ties every field into one working form is document-level and does not get rebuilt, so treat an extracted form page as unreliable to fill in.
Does the extracted file keep the table of contents from the original?
No, bookmarks live in a separate top-level structure that is not tied to individual pages, so a new file built from a handful of extracted pages does not carry that navigation over.
Can I extract the same page into two different output files?
Yes, run the extraction twice with different selections. There is nothing stopping the same source page from appearing in more than one result.
What happens if I select zero pages by mistake?
The extract button stays disabled until at least one page is selected, so there is no way to accidentally generate an empty file.
Does extraction change the rotation of a page I already rotated?
No, rotation is stored on the page itself and is copied along with everything else, so a page you rotated beforehand keeps that rotation in the extracted file.
Want the general split walkthrough, or need to actually divide the file rather than copy pages out of it? See the split PDF hub. Need to cut at precise ranges like 1-5, 6-10, and 11-20 instead of picking individual pages? See split a PDF by page range.
Related tools
- Split PDF. Split a PDF into separate files by page range.
- Remove Pages. Delete unwanted pages from a PDF document.
- Reorder Pages. Drag and drop page thumbnails to change their order.