Split PDF pages made simpler, in your browser

A long PDF can be awkward when different people need different sections. A scan might contain four unrelated forms, a report might have one appendix for each recipient, or a packet might need to be filed as separate documents. This split PDF tool takes one source file and creates new PDFs from its existing pages. You can save every page separately, group pages by a fixed count, name several custom ranges, or extract selected pages into one new document. The source PDF stays unchanged on your device.

The distinction between splitting and extracting is useful. Splitting usually produces several output files from one input. Extracting selected pages produces one output file containing only the page positions you chose, even when those positions are not adjacent. This page offers both workflows with separate controls. It does not divide the visible surface of a single page into halves, crop margins, remove printed page numbers, or rewrite the text inside a page. Its unit of work is the PDF page.

One PDF per page is appropriate when each page is a standalone receipt, certificate, or sheet that needs its own attachment. A four-page source becomes four one-page PDFs. The download is a ZIP because several files have to be delivered together. Each filename includes the source name and its original page position, so the contents of the archive can be sorted and checked after extraction.

Groups of a fixed page count handles a regular sequence. Enter 2 to split an eight-page scan into pages 1–2, 3–4, 5–6, and 7–8. If a nine-page source is grouped by two, the last output contains page 9 alone; the tool does not add a blank sheet. A group size larger than the source page count creates one PDF. This method works well when every record occupies the same number of pages, but inspect the page boundaries first if a scan includes a cover or separator page.

One PDF per custom range gives each specified interval its own file. Type 1-3;5-6;9 to create three PDFs from those three ranges. A semicolon separates output files. Ranges are inclusive, so 1-3 contains source pages 1, 2, and 3. This method can intentionally leave out source page 4 or pages 7–8. The tool rejects an empty interval, an out-of-bounds page, and overlap between intervals. It will not silently place the same source page in two output files. To put nonadjacent pages into a single file, use the extraction method instead.

Selected pages into one PDF accepts entries such as 1-3,7,9. Commas combine page positions into one new file. You may also type 4,1 when you deliberately want source page 4 before source page 1 in the output. The field rejects duplicate page selections, including duplicates hidden by a range such as 1-3,2. One resulting PDF downloads directly rather than inside a ZIP. The filename reports the span and number of selected pages; for nonadjacent selections, inspect the exact order in the downloaded document.

1. Choose one PDF. The browser reads the file locally and checks its page count. The current limit is 30 MB and 150 pages per source. An encrypted, unreadable, or interactive form PDF is rejected instead of producing a misleading result.
2. Select a method. Choose individual pages, a fixed group size, custom intervals, or one extraction. The interface shows only the parameter needed for the chosen method.
3. Check the output plan. Before processing, read the number of planned files, the number of selected source pages, and whether the download will be a PDF or a ZIP. The action button remains unavailable while the page selection is invalid.
4. Run the task. Progress reflects how many output PDFs have been created. Cancel if you chose the wrong method or need to stop the work. After cancellation you can correct the settings and retry.
5. Download and review. The processor reopens each generated PDF to check its page count, dimensions, and rotation against the chosen source pages. Save the result and open it in a PDF reader to inspect important text, images, and document features.

All page numbers in this tool are positions in the input PDF, starting at 1. They are not the numbers printed in a header or footer and are not necessarily the page labels displayed by a specialized reader. A file may call its first PDF page “i,” “Cover,” or “Page 7” visually. Enter 1 to select the first physical PDF page regardless of that printed label. If the count looks surprising, open the source in a reader and count from its first page before entering ranges.

The new filenames start with the source filename and add source page positions. For example, a file named report.pdf split into groups of two can produce report-p1-p2.pdf and report-p3-p4.pdf. An extracted selection uses a label showing its first and last selected positions and the number of pages. These names help trace the output back to the input; they do not add or change visible numbers printed on the pages. When several files are produced, the ZIP is a transport container for those PDFs, not a different PDF format.

The tool copies selected PDF pages rather than rendering every page as an image. This preserves existing page dimensions and rotation and avoids intentional whole-page rasterization. It does not improve a blurry scan or create a text layer for a scanned image. Selectable text in a typical source page is expected to remain selectable, but the source may use uncommon fonts or structures that need manual inspection. Check one text-heavy page, an image page, and any rotated or unusual-size page in the result.

Your document is read through the browser file interface and processed by a worker in the same page. The tool does not send its contents to a file-processing server. A temporary browser download URL is created for the result and revoked when you change settings or reset the tool. The website may still request its own code, styles, and fonts over the network, so local file processing is not a promise that the whole website operates offline.

A source is limited to 30 MB and 150 pages, and generated PDFs are limited to 120 MB in total for one task. These are limits of this browser implementation rather than of the PDF format. Splitting into many small files can require more memory than saving one file because resources used on several pages may be copied into separate outputs. If the task reaches the generated-size limit, split a smaller section at a time. A large but sparse source may produce a small output; an image-heavy source may produce much larger results.

Encrypted PDFs are not decrypted. Interactive forms are stopped because copied pages may appear normal while field behavior or entered values fail in another reader. Document-level bookmarks, internal links, comments, attachments, tagged accessibility structure, and certificate-backed signatures are not guaranteed to survive a page copy. A digital signature may become invalid when a signed document is changed. If one of these features matters, make a suitable reviewed copy with an approved workflow and inspect every relevant output before sharing it.

Choose “Selected pages into one PDF” and type positions such as 2-4,7. The result is one new PDF with those source pages in the order entered. The original document remains unchanged. If you want each chosen interval to become a separate file, use custom ranges with semicolons instead.

Yes. Choose “Groups of a fixed page count” and enter 2. The tool creates consecutive pairs. If the page count is odd, the final output contains one page. Check that any cover page or blank scan does not shift the pairs before you share the ZIP.

The plan reports an error and the task button stays disabled. Custom output ranges may not overlap, and an extraction selection may not repeat a page. Fix the range rather than relying on the tool to guess whether duplication was intentional.

Several output PDFs are packaged in a ZIP so they arrive as one download. Open the archive to access the individually named PDFs. A method that produces only one output, including selected-page extraction, downloads a PDF directly.

This tool copies PDF pages without intentionally turning them into screenshots or downsampling their images. Splitting cannot repair a low-resolution source. Open the result and compare an image and a text page with the source, especially when the document will be printed or submitted.

No. It separates existing pages into files. Cutting or cropping the visible area of one page is a different operation, with different consequences for hidden content outside the crop area.