Keeping only the pages you name
You type the page numbers or ranges you want, and the tool copies exactly those pages into a new PDF. Everything else is left out.
The Pages to Extract box accepts comma-separated values such as 1,3,5-8. A range like 5-8 expands to pages 5, 6, 7 and 8. The result is saved as extracted.pdf, and your original file is never changed.
The output order follows your input, so typing 4,2 produces page 4 first and page 2 second, and typing 2,2 gives two copies of page 2. That makes it handy for quickly repeating a page, such as a blank form, several times in one file. Whitespace around commas and dashes is fine, so 1, 3, 5-8 works the same as 1,3,5-8.
When to extract and when to choose another tool
Extracting is best when you know what you want to keep. If it is easier to name what to remove, use the delete tool.
• Sending a client only the signature page or a single chapter from a long report.
• Pulling the pages you want to run through OCR from a bulky scan.
If you would rather remove a few unwanted pages from a long file, Delete PDF Pages is quicker, since you list only what goes. To break a file into several parts at once, Split PDF fits better.
Practical limits
Selection is typed, not clickable; bad numbers are silently ignored; only one PDF is handled per run.
There is no preview, so you need to know the page numbers in advance. Numbers beyond the last page are ignored without a warning, and a list with no valid numbers yields an empty PDF. Ranges must be written low to high, such as 5-8; a reversed range like 8-5 selects nothing.
Because pages are copied into a fresh document, extras such as bookmarks or document properties from the source may not carry across. Processing is done in your browser, so the file is not uploaded.