Extract PDF Pages
Extract pages from one PDF up to 50 MB into a single new PDF. Enter page numbers such as 1,3,5-7; the output follows your order and is processed in your browser without uploading.
How to use
- Choose your PDF: Drop one PDF onto the box or click to pick it; files up to 50 MB are accepted, one at a time. The page count appears next to the file name, and 1 is filled into the page field for you.
- Type the pages you want: In Pages to extract, separate page numbers with commas and use a hyphen for a range, like 1,3,5-7. Page numbers start at 1, and the new PDF follows the order you type.
- Extract and save the new PDF: Click Extract pages, scroll through the preview of extracted.pdf, then download it. On a phone the save button opens your share sheet when the browser can share files, and downloads the file otherwise.
How to extract pages from a PDF without uploading it
This tool takes one PDF plus a list of page numbers and hands back a single new PDF containing only those pages. Nothing is sent anywhere: the extraction code — the pdf-lib library — is downloaded with the page and runs inside your browser, so your file is read straight from your disk into memory, the new document is assembled on your device, and the result is saved from there. There is no upload, no server copy and no sign-up, which is what makes it usable for contracts, payslips, medical letters and ID scans.
The file and the result only exist in the open tab. Nothing is written to cookies or browser storage, so refreshing or closing the page clears both and there is no history to delete. One PDF is handled at a time, and the drop box is gone once a file is loaded: to work on a different document, remove the current one with the ✕ next to its name first. Only PDFs are accepted, with a 50 MB ceiling per file.
Once you pick a file, the tool reads how many pages the document has and shows it beside the name, and it puts 1 in the field so a single click already produces something. Change that to the pages you actually want. If the text can't be read as page numbers you get an example to copy (1,3,5-7); if a number is higher than the document has, the message names the real page count instead. A damaged file or a password-protected one is reported as such, because the PDF is opened before the page numbers are checked.
The new PDF follows the order you type
Pages come out in the order you typed them, not in the order they sit in the document. Type 5,1 and page 5 lands first; type 1,3,5-7 and you get the original sequence. That makes the tool a quick way to pull pages and re-sequence them in one step — handy when a scanner put a double-sided batch out of order, or when a reviewer only wants the summary page in front of the detail pages.
Two details follow from that. A page number used twice is kept once, so 2,2,3 produces a two-page file. And ranges only run upwards: 7-5 is treated as a formatting mistake rather than a reverse range, so to flip pages around write them out individually as 7,6,5.
If one page number in the list is invalid, the whole run stops and nothing is produced — you won't get a partial file containing "the pages that worked". Fix the number and click again; your original PDF is never modified, so there is nothing to recover.
Extract, split or delete pages — which one you need
Keep a few pages, one file, your order → Extract PDF Pages Turn one PDF into several files → Split PDF Throw a few pages away, keep the rest → Delete PDF Pages
Extracting means naming the pages you want to keep, and everything else is left behind. Deleting pages works the other way round: you name the pages to throw away and the rest stays in its original order. When you want to keep 3 pages out of 40, extracting is far less typing; when you want to drop page 12 from a 40-page report, deleting is.
Splitting is different again, because it produces more than one file — one PDF per page, or one per range you define. This tool always produces exactly one file, no matter how many pages you select, so "each page as its own PDF" is a job for the splitter.
Need one file per page or per chapter instead? Use Split PDF.
What the extracted PDF keeps, and what it loses
Whole pages are copied as they are, so everything printed on them survives: text stays real text that can be selected, copied and searched, images keep their original resolution, vector drawings stay sharp, and each page keeps its size, orientation and rotation. There is no re-layout, no re-compression and no turning pages into pictures, so an A4 scan, a US Letter form and a landscape spreadsheet can all come out of the same document unchanged.
Document-level extras do not travel with the pages. The new file is created empty and only the selected pages are added to it, which means bookmarks (the clickable outline in a viewer's sidebar), attachments and the document-level structure behind fillable forms are not carried over, and the title and author fields come out blank with pdf-lib listed as the producing software. Fill in and save forms before extracting, and rename the downloaded file yourself if the fixed name extracted.pdf doesn't suit you.
Internal links are not rewritten, so a link pointing at a page you didn't select is certain to be dead, and even a link whose target page is in your selection isn't guaranteed to land correctly — especially once you have changed the page order. If links matter in your document, open the download and click a couple of them before sending it on. Encrypted PDFs are a hard stop: the tool can't ask for or remove a password, so unlock the file in a reader first, save an unprotected copy and extract from that. Whatever you feed in, the file you get back is never password-protected.
Why the file may not get much smaller, and what the tool can't do
A few pages taken out of a big PDF usually won't be smaller in proportion to the page count, because the new file still has to carry the shared resources those pages reference — embedded images and fonts come across whole. In our test sample, a 4-page image-heavy PDF of 825 KB still measured 823 KB after extracting a single page, while a 5-page text-only PDF went from 4,052 B down to 1,091 B for one page. These two samples show different size reductions; your result depends on the resources referenced by the selected pages.
Pages are selected by number only. There is no thumbnail picker, and no way to say "odd pages", "every third page" or "every page containing the word invoice" — the field accepts single numbers and upward ranges, nothing else. You also can't extract part of a page (a crop, a table, one photo), process several PDFs in one go, convert the result to Word or images, or push the extracted pages back into another document. The output is always a PDF.
The work happens on the page itself rather than in a background thread, so the button reads Extracting… and is disabled while it runs. For a heavy, image-rich document that can mean a short pause where the tab doesn't respond; it is normal and the file is not being sent anywhere during it.
If the extracted file still has to fit an upload limit, run it through Compress PDF.
Frequently asked questions
- Are my PDFs uploaded to a server?
- No. The PDF is read inside your browser and the new file is built there too, so the document never leaves your device and no copy is stored online. Closing the tab clears the file and the result, because nothing is written to browser storage.
- How do I type the pages I want to extract?
- Use commas for single pages and a hyphen for a run of pages: 1,3,5-7 gives you pages 1, 3, 5, 6 and 7 in one file. Page numbering starts at 1, decimals and 0 are rejected, and a number larger than the document gets a message naming the real page count.
- Can I extract pages in a different order or reverse them?
- Yes — the output follows the order you type, so 5,1 puts page 5 before page 1. Ranges have to run upwards, so 7-5 counts as a formatting mistake; write 7,6,5 to reverse those three pages. Typing the numbers smallest first keeps the original order.
- Why is the extracted PDF almost as big as the original?
- Because the pages you kept bring their shared resources with them — embedded images and fonts are copied whole rather than trimmed to the pages that use them. Our 825 KB, 4-page image-heavy sample still came to 823 KB after extracting one page. Our 5-page text-only sample went from 4,052 B to 1,091 B when page 1 was extracted; other documents may differ.
- Do bookmarks, links and the document title survive?
- Bookmarks, attachments and the title and author fields do not: the new PDF is created fresh with only your chosen pages added. Internal links are not re-pointed either, so links to pages you left out are dead and the rest aren't guaranteed. Page content itself is untouched. Visible URL text is copied with the page; check any clickable links in the downloaded PDF.
- Can I extract pages from a password-protected PDF?
- No. Encrypted files are rejected as soon as they are opened, and the tool can't prompt for or strip a password. Open the PDF in a reader with its password, save or print an unprotected copy, and extract from that copy instead.
- Can I get each page as a separate PDF file?
- Not here — every page you select is merged into one file called extracted.pdf. Use the Split PDF tool when you need one PDF per page or per range, then extract or recombine from those parts if you still need to.
- Does it work on iPhone and Android?
- Yes, in any modern mobile browser. Pick the PDF from Files, Google Drive or your downloads folder, type the page numbers, and tap the save button: it opens your share sheet when the browser supports sharing files, and falls back to a normal download when it doesn't.