Split PDF Online

Break a multi-page PDF into one file per page, or pull out the exact ranges you need.

In your browser—your files never leave your device.

Files never leave your browser. Splitting happens 100% locally on your device.

Splitting a PDF is the cleanup step when somebody has merged twelve unrelated things into one document and now you need to extract just the parts you care about. This page does it two ways: either output one PDF per page (good when you want to keep some pages and throw out others), or pull out specific ranges like 1-3, 5, 8-12 in a single pass. Both modes run in your browser via pdf-lib, so the source file never uploads. The page count is detected when you drop the file in, so you can confirm the document loaded correctly before splitting. There is no signup, no daily cap, and no watermark on the outputs. The how-to below covers both modes and the FAQ answers the questions that come up after the first successful split — especially around encrypted PDFs, file-size ceilings, and what to do when the input is too big for browser memory.

How to split pdf online

Drop in your PDF

Single file. The page count appears immediately so you can confirm you uploaded the right document and know the valid range for the range field.
Pick a mode

"Each page" outputs one PDF per page. "Custom ranges" lets you type something like "1-3, 5, 8-12" to get those pages bundled into one output PDF.
Click Split

The page builds the output PDFs in memory using pdf-lib. A 50-page split-each operation takes 5-10 seconds on a modern laptop.
Download the outputs

Each output appears as its own download link, named with a 3-digit page number so re-assembling later in the right order is trivial.

Why use this tool

You scanned both sides of every page of a 40-page document and now you have 80 page-images, but you only need pages 12-18 to email to someone. You downloaded a journal article bundled with five other articles in one PDF — you want only your article as a clean file. You have a year-end statement that combines monthly summaries, and you want each month as a separate file so you can drop them into the matching folders. You signed a multi-section contract and your lawyer wants only the appendix. You are studying from a textbook PDF and want to pull out individual chapters for offline reading on your phone. Splitting per-page is the brute-force approach; splitting by ranges is the precise one. Both are here, both run locally, neither needs a signup. The split tool pairs naturally with the merge tool — split, throw out what you do not need, merge what is left in the order you want.

Features

Two modes: each-page and range

Each-page splits the input into N separate PDFs, one per page, named page_001.pdf through page_NNN.pdf. Range mode parses "1-3, 5, 8-12" syntax and gives you one PDF containing exactly those pages in that order. Most other split tools only do one or the other, so having both means you do not need to round-trip through a second tool for a common workflow. The range parser validates against the page count, so typos surface as friendly errors rather than producing silently wrong output.

Zero-padded numbering keeps the order

Output filenames use a 3-digit page number (page_001, page_002, ... page_017), so when you drop them into a folder they sort naturally instead of going 1, 10, 11, 12, 2, 3 in the way that default Finder and Explorer sorts treat numbers as text. This sounds obvious but several other split tools get it wrong and produce a mess that is annoying to recombine. Re-merging in the right order with the merge tool is then trivial because the alphabetical sort matches the page order.

Page count shown before you split

As soon as you drop in the PDF, the page reads the document structure with pdf-lib and shows you the total page count. So if you uploaded the wrong file or your PDF is actually one page (which cannot be meaningfully split), you find out instantly instead of after waiting 30 seconds for the split to fail. The page count also doubles as a sanity check on the range field — you cannot type "1-100" on a 30-page PDF and have it silently produce something weird.

Runs in your browser

Splitting uses pdf-lib in a Web Worker — the source file never leaves your device. The output PDFs are generated locally and downloaded directly from in-memory blob URLs that get revoked when you close the tab. Verify this in the Network tab if you are skeptical. For sensitive documents (medical, legal, financial) this is the only safe option short of installing desktop software like Acrobat or PDFsam.

Privacy & security

Splitting a PDF is one of the few operations that runs entirely on metadata — pdf-lib copies the page objects you select into a new document inside your browser tab and writes the result back to disk. No upload, no server-side worker, no temp files anywhere but your browser's memory. The page-range parser that translates '1-5, 8, 11-13' into actual page indices is plain JavaScript running on your machine. If you split a 200-page PDF into 200 separate files, that is 200 downloads triggered by your browser, not 200 server responses.

Frequently asked questions

What if my PDF only has one page?

Then there is nothing to split. The page tells you the page count immediately when you upload, so you will see it is 1 right away. If you wanted to extract a region from a single-page PDF (e.g. just the top half) you need a different tool — image cropping after PDF-to-JPG conversion, or a proper PDF editor like Acrobat. The split tool only operates at page boundaries because that is what the PDF format natively supports.

Can I split by custom page ranges?

Yes. Pick the "Custom ranges" mode and type something like "1-3, 5, 8-12". You get a single output PDF containing exactly those pages in that order. Multiple ranges separated by commas are combined into one output — for separate outputs per range, run the split multiple times changing the range each time, or use each-page mode and pick the files you want afterward. The range parser ignores spaces around commas and dashes, so "1-3,5,8-12" works the same as "1 - 3, 5, 8 - 12".

Are my files uploaded anywhere?

No. The split runs entirely in your browser via pdf-lib. The source PDF stays on your device, and the output PDFs are generated locally. The Network tab of your browser's developer tools will confirm there are no outgoing requests for the file content during the split. The page itself loads from our CDN once when you visit; after that, all operations are local.

What happens with encrypted PDFs?

The page detects encryption when it tries to load the document and surfaces an error. You need to unlock the PDF first (with the PDF unlock tool — you must know the password), then split the unlocked copy. The split tool itself does not handle password input because conflating the two would let someone build a brute-force tool out of it, which is not the intent. The two-step flow keeps the tools focused.

How big a PDF can I split?

A few hundred megabytes is fine on a typical laptop. The bottleneck is browser memory because pdf-lib loads the entire document into RAM. For files in the 500MB+ range you will start hitting memory limits, especially on a phone or Chromebook with limited RAM. If your input is that big, compress it first with the compress tool to reduce the memory footprint, or split it in two passes — pages 1-N, then pages N+1 to end — using the range mode.

Will form fields survive the split?

Mostly yes. pdf-lib preserves form-field annotations when copying pages, so a single split page with a form on it will still have the form widgets in the output. There are edge cases — if a single form field spans multiple pages (rare but possible) it may not render correctly after splitting because the underlying field object is shared. For typical one-page form sections, the split is clean.

Can I drag pages to reorder before splitting?

Not in this version. The split mode just splits — for reordering you can use the merge tool after splitting (each page becomes its own file, then merge them in any order). Adding drag-to-reorder directly here is on the list, but every UI complication adds load time so we are picky about what gets added. For now, split + merge is the two-step workflow for arbitrary reordering.

What about bookmarks and the table of contents?

Bookmarks (the navigation tree in the left pane of most PDF readers) are tied to specific pages in the source document. When you split into individual pages, each output PDF gets only the bookmarks that pointed to that page. When you use range mode, the output PDF gets the bookmarks that fell within the range, with their page numbers adjusted to the new positions. It usually works for typical reports; very nested or cross-referencing bookmarks can lose context.

Will the output PDFs be smaller than the input?

Each individual output is smaller because it has fewer pages, but the sum of all outputs is roughly the same size as the input. pdf-lib does not re-encode page content during the split, just copies it. So the per-page filesize matches what that page weighed in the original PDF. If your original was 20MB across 100 pages, each split page averages 200KB.

Does it work on a phone?

Yes, in any modern mobile browser. The split is light on CPU but uses a fair amount of memory because pdf-lib loads the whole document at once. A modern phone can comfortably split documents up to about 100MB; older phones with limited RAM will hit ceilings earlier. The download links work natively on iOS and Android — tapping each one saves the file to your Downloads or to wherever your browser saves files.

What does the range syntax actually accept?

Comma-separated ranges where each range is either a single page number ("5") or a start-end range ("3-7"). Spaces around commas and dashes are ignored, so "1-3, 5, 8-12" works the same as "1-3,5,8-12". The parser does not accept open-ended ranges ("-5" or "5-"), step values ("1-10:2"), or reverse ranges ("10-1"). If you type something invalid, the tool surfaces a clear error rather than silently producing wrong output.

Can I get a single PDF with non-consecutive pages?

Yes — that is exactly what range mode does. Type "1-3, 5, 8-12" and you get one output PDF containing pages 1, 2, 3, 5, 8, 9, 10, 11, 12 in that order. They are bundled into a single file, not nine separate files. This is useful when you want to keep specific sections of a long document — say, pages 1-3 (cover + intro), 5 (a summary), and 8-12 (appendix), discarding the rest.

Will splitting change the file size?

Each output PDF is smaller than the input because it has fewer pages, but the sum of all per-page outputs is roughly the same as the input — pdf-lib does not re-encode page content during the split. Each output also has a small fixed overhead (PDF header, cross-reference table, document structure) of around 2-5KB per file, so 50 tiny single-page outputs end up totaling slightly more bytes than the source 50-page document. For real-world documents this overhead is negligible.

About this tool