Split PDF Into Parts

What is PDF splitting?

Splitting PDF (split PDF) is creating multiple new PDF files from a single source. Depending on the chosen mode, the result can be one file (with a subset of pages), several files in a ZIP archive, or one file with pages removed. Page quality, fonts, images, and embedded objects are preserved unchanged - our service operates at the PDF object model level.

Splitting is useful when you need to extract several pages from a large document, remove unnecessary pages, break a report into chapters, or extract appendices as separate files. One of the most popular requests on any PDF service - this functionality is widely used in work and study.

4 splitting modes

Our service offers 4 different modes, each suited to its task. Selection is one click on the mode card, after which the corresponding configuration UI is shown.

Mode 1: Extract pages

Specific pages are selected from the source PDF - the result is one new PDF containing only those pages in the specified order. Enter numbers separated by commas, ranges with hyphens, use the keyword end for the last page.

Examples:

• 1, 3, 5-7 - extracts pages 1, 3, 5, 6, 7
• 1-end - extracts everything (valid but pointless)
• 10-end - extracts pages 10 to last

Result: a single PDF.

Most common scenario: pulling an introduction from a contract, a chapter from a book, specific pages from a report.

Mode 2: Remove pages

You select pages to remove - the result is one PDF without those pages. The opposite of extraction: mark what you don't need.

Examples:

• 2, 4 - removes the second and fourth pages (output without them)
• 1-3 - removes the first three pages (cover, abstract, table of contents)
• 15-end - removes the document's tail

Result: a single PDF.

Useful when removing a blank page, a watermarked page, ads, or a cover.

Mode 3: By ranges

Each comma-separated range becomes a separate PDF - the result is a ZIP archive with N files.

Examples:

• 1-3, 4-6, 7-end - three separate PDFs in ZIP: pages 1-3, 4-6, 7-end
• 1-10, 11-20, 21-30 - three volumes of 10 pages each

Result: ZIP.

Used when a document needs to be broken into chapters or volumes: a book into chapters, an annual report into quarters, a contract into main sections.

Mode 4: By pages

Each page of the source PDF becomes a separate file - the result is a ZIP with N files (one per page). No input required - just "split into singles".

Use cases:

• Scan of a 50-page book → 50 separate PDFs for archiving
• Presentation → each slide as a separate file for distribution

Result: ZIP.

This is the "brute force" mode: everything into separate files, no extra parameters.

Visual page selection

The main feature of our service - client-side PDF rendering via the pdf.js library (by Mozilla). After upload, the file is rendered in your browser, pages are displayed as a grid of thumbnail images. You click pages you want to select (for extraction or removal) - selected ones are highlighted in color (green for extraction, red for removal).

Client-side rendering benefits

• Speed - no round-trip to server for previews; rendering happens immediately on upload.
• Privacy - the file is rendered in your browser, not sent to server until you click "Split".
• Lower server load - previews don't take space in our storage.
• Works offline - if the connection drops, you can browse previews and change selection.

Dual selection

Besides clicking thumbnails, you can type page numbers in a dedicated text field. Thumbnail grid and text field are synchronized: clicking thumbnails updates text, typing text highlights corresponding thumbnails. Convenient for 200-page PDFs - scrolling the grid is slow, easier to type 42, 78, 95-100.

How it works on the server

Extract and remove

Under the hood - qpdf, an open-source C++ tool for working with PDF. A command like qpdf input.pdf --pages input.pdf 1,3,5-7 -- output.pdf creates a new PDF from only the specified pages. For "remove" mode we compute the inverse: if the document has 10 pages and pages 2,4 should be removed - qpdf gets spec 1,3,5-10.

By ranges

Each range runs a separate qpdf invocation, results are zipped via the standard Rust zip library. Filenames inside ZIP: <task_id>_range_001.pdf, _range_002.pdf, etc.

By pages

qpdf supports a special mode --split-pages=1 - each page goes into a separate file. Server collects all into a ZIP.

Performance

• Extracting/removing 10 pages from a 100-page PDF - usually 1-2 seconds.
• Splitting into 50 files and packaging into ZIP - 5-10 seconds.
• Execution time limit - 3 minutes, after which the task is cancelled.

Technical details

Range specification

Compatible with qpdf page-range syntax:

• Numbers: 1, 5, 10
• Ranges: 1-3, 5-end
• Comma-separated lists: 1, 3, 5-7, 10-end
• end or z - the last page

What is preserved

• Text, fonts, images, hyperlinks within a page.
• Document metadata (author, title, creation date) - usually copied to the result.
• Bookmarks pointing to preserved pages - work.

What may be lost

• Bookmarks pointing to removed pages - become "dead" links.
• Digital signatures - valid only for the original document; become invalid after split.
• Interactive forms - fields are preserved but may not work correctly if the form referenced pages outside the result.

Limits

The 200-page preview limit on free tiers is a technical guard against browser overload. Paid tiers render without limits.