Extracting pages from a PDF means pulling specific pages out of a larger document and saving them as a new, standalone file. The original document stays untouched. This is the tool to reach for when you need just chapter three of a textbook, the signature page from a contract, or slides 10 through 25 from a presentation deck. Everything happens in your browser — no uploads, no accounts, no waiting.

When to use this tool

The clearest use case is isolating a section of a long document. You received a 90-page policy manual and only need pages 14 through 22 for a meeting. Extracting those nine pages gives you a focused file you can share without forcing colleagues to scroll through material they do not need.

Another common scenario is pulling a single page for a form submission. Government agencies and landlords often ask for "page 2 of your bank statement" or "the declarations page of your insurance policy." Rather than sending the entire document — with all your personal information — you extract the one page they actually need.

Research and academic work benefit from extraction too. Grab the bibliography, the methodology section, or a specific figure from a published paper. Keep the relevant pages together in a file you can annotate and reference without hauling around the full publication.

Extraction also serves as a building block for more complex workflows. Extract pages from several source documents, then merge the results into a custom compilation. This is how people assemble training packets, course readers, and proposal appendices from scattered source files.

How page extraction works inside a PDF

A PDF stores each page as an independent object in the file's internal structure. Page objects reference their own content streams — the instructions that draw text, shapes, and images on the page. They also reference shared resources: fonts, color profiles, and embedded images that may be used by multiple pages throughout the document.

Extraction starts by identifying the page objects that correspond to the numbers you specified. The tool then traces every reference from those page objects — content streams, font dictionaries, image objects, annotation dictionaries — and collects all of them into a dependency set.

Here is where it gets interesting. PDFs are aggressive about sharing. A single font object might be referenced by every page in the document. An image used as a watermark appears once in the file and is referenced by dozens of pages. When you extract pages, the tool cannot just take the page objects in isolation. It must follow every reference chain and pull in every resource that the selected pages depend on.

This is why extracted PDFs are sometimes larger than you would expect. If you extract one page from a 50-page document, and that page uses a font that is embedded once for the entire document, the extracted file must include the full font embedding. The same applies to shared images, ICC color profiles, and embedded file attachments. The new document needs to be self-contained — it cannot point back to the original file for missing resources.

After collecting all necessary objects, the tool writes a new PDF with a fresh cross-reference table and its own page tree. The result is a valid, independent PDF that has no dependency on the source file. Object numbers are reassigned, the page tree is rebuilt, and any metadata from the original document (author name, creation date, producer software) can optionally be carried forward or stripped.

One subtlety worth mentioning: cross-references between pages can break during extraction. If page 5 contains a link that says "see page 12," and you only extract pages 5 through 8, that link will have nowhere to go. Internal links targeting pages within your extracted range are preserved and retargeted to the correct position. Links pointing outside the range become dead.

Common issues

File size seems disproportionately large. This is the shared resource problem described above. Extracting five pages from a 200-page document does not give you a file that is 2.5% of the original size. The extracted file carries all fonts, images, and resources those five pages need, even if those resources were originally shared with the other 195 pages. For documents with many embedded fonts or high-resolution images, the overhead can be significant. There is nothing wrong with the extraction — the new file simply has to be self-sufficient.

Fonts look different or fall back to substitutes. Occasionally, font subsetting interacts poorly with extraction. The original PDF may contain a font subset that only includes the glyphs used across the entire document. When that subset is pulled into the extracted file, it should contain every glyph the extracted pages need. In rare cases with unusual font configurations, a viewer might display a substitution. This is uncommon with modern PDFs but can happen with older files.

Cross-reference links break. Internal navigation — links in a table of contents, "see page X" references, linked footnotes — only works within the extracted page set. Anything pointing to a page that did not make it into the new document will either do nothing when clicked or display an error, depending on the reader. External URLs (links to websites) always work regardless.

Bookmarks point to missing pages. If the original document had bookmarks and you extract a subset that does not include every bookmarked page, those dangling bookmarks are removed. Bookmarks that target pages within your extracted set are preserved and renumbered.

What to expect from our tool

Load your PDF and you will see a page-by-page thumbnail grid. Click individual pages to select them, or type a range expression in the input field. The range syntax supports commas, dashes, and mixed formats: entering "1, 4-8, 15" selects exactly those pages. You can specify pages in any order. The output file will match the sequence you define, which means extraction doubles as a lightweight reorder operation.

Once you confirm your selection, the tool assembles the new file in your browser. Small extractions finish nearly instantly. Pulling 20 pages from a large image-heavy document might take a few seconds while the browser processes the shared resources. No data is sent anywhere during this process.

The download gives you a new PDF. Open it in any reader — Adobe Acrobat, Preview, Chrome's built-in viewer, Foxit, Sumatra. The file conforms to the PDF specification and works everywhere the original did. Your source document is completely unchanged. Run the extraction again with different pages if you need another subset.