Keep the PDF pages that matter

Start with the right document

Use this when you need a chapter, a few meeting notes or an appendix from a longer PDF. The PDF page extractor creates a new document from your selected pages. It does not edit the original, compress images, remove information inside a page or perform OCR.

You need an unencrypted PDF under 25 MB with no more than 500 pages. Keep the source somewhere safe. If it has a digital signature, extraction creates a different document: the original signature will not authenticate the new file.

For a safe first run, download our five-page practice PDF. It contains clearly labeled sample pages and no private information.

Identify physical page positions

A report can display “i, ii, iii, 1, 2…” even though a viewer counts the first page as 1. This tool uses that physical count. Open your PDF and use the viewer’s page-position indicator, not a number printed at the bottom of the page.

Suppose a cover occupies physical page 1, and the first page of the report occupies physical page 2. Entering 1 extracts the cover. Entering 2 extracts the first report page.

Write down the positions you need before starting. This avoids discovering after a download that a title page changed your numbering.

Extract a specific selection

1. Open PDF page extractor.
2. Select Choose a PDF and pick your source file. You can also use Load sample.
3. Enter your selection in Pages to keep. Use commas between separate positions and a hyphen for an ascending range.
4. Select Extract pages. If a position falls outside the document, correct it using the page count in the error message.
5. Read the selection summary, then choose Download result.

On the five-page sample, try:

2-4, 1

The output should contain physical pages 2, 3, 4, 1, in that order. Selection order is preserved. Repeated pages are included only once, so 1-3, 3 exports three pages. A descending range such as 4-2 is not supported; use 4, 3, 2 if you need that order.

Check the exported copy

Open the downloaded PDF in a viewer. Confirm its page count, first page, last page and orientation. Look for clipped content, missing images and unexpectedly blank pages. If you rely on clickable links, test them individually; links pointing to an omitted page may no longer work as intended.

Document-level bookmarks, signatures and interactive form behavior are not guaranteed to survive. This tool copies pages rather than rebuilding the original document’s navigation and form structure. For a signed or interactive filing, use the original or a workflow approved by the receiving organization.

Rename the output to explain its scope, such as project-notes-pages-2-4.pdf. Keep the source alongside it until the recipient confirms that the copy meets their needs.

When something goes wrong

• Encrypted or unreadable PDF: open it in a trusted viewer. If you are authorized to create an unencrypted copy, export that copy there and retry. This tool does not remove passwords.
• Page outside the document: count physical positions again. The source may contain fewer pages than its printed page labels suggest.
• A large or slow file: choose a smaller source. Processing uses your browser’s memory; phones can run out of memory before the file-size limit is reached.
• Sensitive content remains: page extraction only excludes whole pages. It is not redaction. Never cover private text with a rectangle and assume it is removed.

For the final handoff, follow Prepare a PDF for sharing.