Split PDF

What "splitting" a PDF actually means

"Split PDF" is one of those phrases that means three different things depending on who's asking. Some people want to chop a 200-page document into 200 single-page files. Some people want to pull out chapter 4 and email just that. Some people want one PDF cut roughly in half. They're all reasonable, and they're all different jobs.

Our Split PDF tool does the second one: you tell it which pages you want, and it gives you a single new PDF containing exactly those pages, in exactly the order you asked for. No surprise renaming, no zip file of 47 little PDFs, no marketing modal asking you to upgrade. Type "3-7,10,15-20" and download.

If you want every page as its own file, that's coming. If you want to merge two PDFs together, we have a PDF Merger for that. If the question in your head is "remove these pages, keep the rest," you want Delete PDF Pages — same engine, opposite question.

How the range syntax works

The tool takes one input: a list of page numbers and ranges, comma-separated. Spaces are ignored. Pages start at 1, not 0. That's the whole grammar.

The last point is the one most other tools get wrong. iLovePDF and SmallPDF interpret your selection as a set and emit pages in their original order. Ours doesn't. If you type "5,3,1" we give you a three-page PDF with page 5 on top. This is deliberate, and it's a small reordering tool hiding inside a splitter. Need to flip a five-page document? 5-1 expands to 5,4,3,2,1 and you're done.

Out-of-bounds pages produce a clean error instead of silent truncation. If you type 1-50 on a 30-page PDF, you get told "page 50 doesn't exist" — not a half-broken output you discover later.

How to use the Split PDF tool

1. Drop or pick your PDF. Up to 100 MB. The tool reads the file in the browser and shows the total page count next to the file name.
2. Type the page range you want in the input field. The placeholder shows an example for your file's size, so you have something to edit rather than a blank box.
3. Click "Extract pages." A new PDF is built in your browser, in the order you specified.
4. Click Download. The output file is named after your input plus the page range — contract.pdf with range 3-7 becomes contract-pages-3-7.pdf.

There's no Calculate-style button to retry, no "Continue" modal, no email gate. Your PDF never leaves the browser tab. Watch the network panel while you split if you don't believe us — there are zero outbound requests during the operation.

A worked example

Say you've got a 30-page board pack named q3-board-deck.pdf and you want to send only the financials (pages 8–14), the headline customer logo slide (page 3), and the appendix (pages 25–30) to a director who asked for "the numbers and the appendix."

Your range expression: 3, 8-14, 25-30

That's 1 + 7 + 6 = 14 pages. The output PDF has them in the order you wrote them: logo slide first, then financials, then appendix. If you'd rather have the financials first because that's what they actually asked for, swap to 8-14, 3, 25-30. Same pages, different sequence.

Now imagine the director writes back: "Actually I just want the last six pages in reverse order — newest summary first." Range expression: 30-25. Done. That's the whole operation.

Why client-side matters here

The two biggest free PDF tools — iLovePDF and SmallPDF — upload your file to their servers, run the split there, and let you download the result. That works, but it has consequences most users don't think about until later:

• Your PDF is on their disk. Their privacy policies promise quick deletion, and most of the time that's probably true. But you signed a contract is now sitting in some object storage bucket for a few hours, alongside whoever else is splitting a PDF that day.
• 5 MB free-tier caps. SmallPDF's free split capped at 5 MB last we checked; iLovePDF caps at 25 MB unless you log in. Any real document — a scan, a slide deck with screenshots, a contract with embedded images — blows past those numbers immediately. The "upgrade to Pro" wall is the entire business model.
• Watermarks. Some "free" splitters return a PDF with a "Created with Free PDF Tool" footer on every page. Ours doesn't. Your output is byte-for-byte the pages you selected.
• Round-trip time. Uploading a 50 MB PDF on home internet, waiting for a server, downloading it back — that's 30+ seconds of staring at a progress bar for an operation that takes about 200 ms when done locally.

We run the split with pdf-lib, a JavaScript PDF library that reads and writes PDF byte streams directly in the browser. Page operations are mostly metadata work — copying page objects from the input to a new output document — so there's no re-rendering, no quality loss, no font substitution. The output is the same pages, just selected and reordered.

Split PDF vs the alternatives

Quick comparison of how the common free options handle a typical "extract pages 3 through 7 from a 12 MB contract" job:

PDF24 is the only other one with a real client-side mode, and the UX is a maze of settings and ad banners. iLovePDF is the cleanest paid experience if you don't mind uploading. Adobe is fine if you already have a subscription. Everyone else is a usage-throttled funnel toward a Pro tier.

Edge cases worth knowing about

A few situations where Split PDF behaves in specific ways that aren't obvious:

• Bookmarks. Top-level outline entries (the sidebar of bookmarks Acrobat shows) reference page indices that change after splitting. They don't transfer. On-page hyperlinks survive.
• Form fields. Visible form fields copy through with their current values. The interactive behaviour (linked dropdowns across pages, complex JavaScript validation) sometimes survives and sometimes doesn't, because PDF forms are an unfixed swamp. For anything contractual, flatten the form first with our PDF Flatten Form tool, then split.
• Encrypted PDFs. pdf-lib refuses to open password-protected PDFs. Unlock first using macOS Preview (File → Export → uncheck Encrypt) or Adobe Acrobat. We deliberately don't bundle a password cracker.
• Scanned PDFs. A 50-page scan is often 200+ MB because each page is a high-resolution image. Split works fine but the output file size scales with how many pages you keep. Run the output through our Compress PDF if size matters.
• Files over 100 MB. The browser has to load the whole PDF into memory before splitting. On a phone or older laptop, 100+ MB is asking for trouble. We cap there to fail clearly rather than crash silently.

Related PDF tools

Split PDF is part of a small family of page-level operations that all share the same client-side pdf-lib engine and the same no-upload promise:

• Delete PDF Pages — the inverse question. Tell it which pages to drop instead of which to keep.
• Reorder PDF Pages — keep all pages but rearrange the order. Useful when "type 5,3,1" inside Split PDF is more rearranging than splitting.
• PDF Merger — the reverse of splitting. Join two or more PDFs into one.
• Rotate PDF — for sideways scans. Often used together with Split when a scanner produced one wrong-rotation page.
• PDF Page Count — quick page count for a PDF without opening it. Helpful when you're writing your range expression.

Frequently asked questions

Is my PDF really not uploaded?

Yes. The split runs entirely in your browser through pdf-lib. Your PDF bytes travel from your file system into the browser tab's memory, get copied into a new PDF, and come back out as a download. Nothing crosses the network. Open your browser's developer tools, click the Network tab, and run a split — you'll see zero outbound requests during the operation. The Microapp page itself loads from our CDN, like any web page; everything after that is local.

How does the page range syntax work?

Comma-separated page numbers and ranges. 3-7 means pages 3 through 7. 1,5,9 means just pages 1, 5, and 9. 1-3,8,12-15 combines them. Spaces are ignored. Pages are 1-indexed, so the first page is 1, not 0. The output PDF is built in the order you typed, which means 5,3,1 gives you page 5 first, then 3, then 1 — a small reordering feature hidden inside the splitter.

Can I extract every page as its own separate PDF?

Not in one click yet. Each run produces one output PDF. To get N separate files for an N-page document, run the tool N times with 1, 2, 3 and so on. We know that's tedious for large documents; a "burst mode" that produces a zip of single-page PDFs is on the roadmap. If you need it now, ping us.

What's the maximum file size?

100 MB per PDF. The browser has to load the whole file into memory to copy pages, and on phones or older laptops, much past 100 MB starts hitting browser memory limits and crashing the tab. For multi-hundred-MB scans or archives, work in chunks (split into pieces, then split each piece) or use a desktop tool like macOS Preview or Adobe Acrobat.

Does the output keep bookmarks, hyperlinks, and form fields?

Partially. pdf-lib's copyPages preserves the page content, on-page annotations, hyperlinks that point within the page, and the visual appearance of form fields with their current values. Top-level bookmarks (the outline pane in Acrobat) reference page indices that shift after splitting, so they don't transfer. Cross-page links to a deleted page become dead. For anything signature-related, run the source through PDF Flatten Form first.

What about password-protected PDFs?

pdf-lib won't open them. You'll see a "this PDF is encrypted" error. Unlock first: in macOS Preview, File → Export → uncheck Encrypt; in Adobe Acrobat, File → Properties → Security → set to No Security and save. We deliberately don't bundle a password remover — if you don't have the password, you shouldn't be splitting the file.

Can I split scanned PDFs?

Yes. Split treats every page identically whether the page is vector text, embedded images, or a scanned bitmap. The output keeps the original scan quality. Scanned PDFs are often large because the pages are high-resolution images — if your output is bigger than expected, run it through Compress PDF afterwards.

What if I just want to remove a few pages, not extract them?

Use Delete PDF Pages instead. It's the same engine, but the input is "which pages to remove" rather than "which pages to keep." For a 50-page document where you want everything except page 50, "delete 50" is easier to type than "extract 1-49." Use whichever question matches what's in your head.