How to Split a PDF

Splitting means turning one PDF into several output files. The source stays untouched on your device; you get a set of new files, each holding the pages you asked for. The job comes in four shapes, and the difference between them is entirely about where the cut lines go: you draw them by page numbers, by a fixed page count, by the document’s own bookmarks, or by text that marks a boundary. This guide picks the tool for each shape and walks the steps. It belongs to the merge and split family, and the mirror-image job is How to Merge PDF Files.

Which tool should I use to split?

Split PDF covers ranges and fixed counts; bookmarks and keywords get their own tools. The plain split tool handles every cut you can express as page numbers. The two special shapes need dedicated tools because the cut lines come from the document itself:

Split shape	The tool for it	Where it runs
By page ranges (pages 1–3, 5, 8–12)	split a PDF	Browser, free, no upload
Every N pages, one file per page, or N equal parts	split a PDF	Browser, free, no upload
At each bookmark or heading, into chapters	split a PDF by bookmarks	Browser, free, no upload
Where a keyword or pattern appears (one invoice per file)	split a PDF by text	Browser, free, no upload

Everything on this table runs in your browser with no account and no upload. Two neighboring tools finish the family’s work: auto-remove blank pages cleans duplex-scan separators, and rearrange PDF pages handles the case where you want to keep some pages rather than split.

How do I split a PDF, step by step?

1. Open the split tool. Go to split a PDF — no account, nothing uploaded.
2. Load one PDF. The source file opens into local memory only. Nothing is transmitted, and closing the tab clears it.
3. Pick the split mode. Choose by range, every N pages, one file per page, or split into N parts. The page-preview shows how many files the current settings will produce before you commit — check that number against what you expect.
4. Run the split. The tool writes one output file per range or chunk, again entirely in the browser.
5. Download the results. Take the files one by one, or download them all as a single ZIP archive — one download instead of a dozen, which is the practical choice for any split into more than a handful of parts.

A concrete example: a 40-page supplier contract appendix that must go to four reviewers, one section each. Split by range into four parts, download the ZIP, and send each reviewer one file containing only their section. The original 40-page file never left your machine.

How do I split a document that has no page numbers I can rely on?

Bound bundles are the common hard case: a year of invoices in one download, one report per chapter in a single export, a workshop packet where each section starts on an unexpected page. Counting page ranges by hand is exactly the work the two special split tools remove.

• By bookmarks. If the document has an outline — the bookmark pane a reader shows on the left — split a PDF by bookmarks cuts at each bookmark and produces one file per chapter. If there is no outline, the same tool detects heading text on the pages and offers to cut there. It is the right tool for reports, books, and anything with a visible structure.
• By keyword. When each unit starts with a known phrase — “INVOICE”, “Purchase Order”, a case reference, a date stamp — split a PDF by text searches for the word or pattern and starts a new file at each hit. This is the tool for a bundled invoice download: it turns one fat PDF into one file per invoice, which is the shape your accounting software actually wants. It needs a real text layer, so run OCR a PDF first on scans.

Both need the document to carry the markers: bookmarks, headings, or searchable text. A bundle with none of those has to be split by page numbers.

How do I check the split worked?

Before you send or file the outputs, verify three things:

1. File count. The preview told you how many files to expect; confirm you have that many.
2. Page coverage. Open the first and the last output and confirm the pages you wanted are inside, and that nothing between your cut lines landed in two files. A range overlap shows up immediately as duplicated pages.
3. Text layer. If the source was searchable, spot-check that the outputs are searchable too — splitting copies pages, so the text travels with them, but a source that was a scan stays a scan.

For bundles that must also become data, the next step is conversion: How to Convert PDF Files covers the direction you need, including the invoice-to-spreadsheet path.

How do I keep the split outputs traceable?

A split that produces ten or twenty files creates its own problem: which part is which. The tool handles the naming for you — outputs are named after the source file with a part number appended (report-part1.pdf, report-part2.pdf, and so on, in order), the visual page-selection mode names its single output report-selected.pdf, and the download-all archive is always split-pdfs.zip. Three habits close the rest of the loop:

1. Use the ZIP download for anything above a few parts. The archive preserves the part order in its filenames, so extracting it gives you parts in sequence rather than a set of downloads in whatever order the browser finished them.
2. Number the source before splitting when it has no markers. Adding page numbers to the source first means every output carries its original page position, so a part can be cited back to the page range it came from even after the file is renamed.
3. Keep the source until the outputs are verified. Splitting never modifies the source, so the original is the reference copy if a boundary lands in the wrong place.

What can splitting not do?

• It cannot cut inside a page. Every cut lands on a page boundary. If one page holds two invoices, no split tool can separate them — that is an extraction job, not a split.
• It cannot read a scan. Keyword and bookmark splitting need a text layer; a scan has none. OCR the file first, or split by page numbers.
• It does not reorganize. Splitting never reorders, rotates, or deletes. Do that organizing pass first if the source needs it.
• It does not shrink the outputs. The parts together are about the size of the source. If each part must clear a size cap, follow with compress a PDF.

Frequently Asked Questions

Do I need an account to split a PDF? No. Splitting runs in your browser, free, with no upload and no account. Nothing leaves your device.

Does splitting damage the original? No. The source file is read into memory and the outputs are new files. The original is never modified — you can re-split it with different settings as many times as you like.

Why did my keyword split miss a boundary? Usually the marker differs from the pattern: a different word, extra spacing, or a scanned page with no text layer. Open the pages around the missed cut and compare the actual text with your pattern; broaden the pattern or OCR the file. The tool reports where it found each boundary, so the list of hits shows the gaps directly.

Can I split out just one part and keep it? Yes — use the visual page-selection mode rather than a range split. You pick the pages on the preview and the tool outputs a single file, named <source>-selected.pdf, holding only the pages you picked. For a part that is most of the document, split out that part and delete it from a copy with rearrange PDF pages instead.

Can I split a PDF that is too big for the browser? The browser path is bounded by the tab’s memory, and the sizing rules are on the file limits reference page. For a bundle far above that ceiling, the cloud large-file entry on the tool page is the alternative.

---

Splitting is the same simple act four ways: decide where the cut lines come from — numbers, structure, or text — then let the tool draw them. Keep the source untouched, verify the file count and the edge pages, and you have a set of files in the shape the next step actually needs.

Do It Directly

Ready to cut up the file in front of you? Split a PDF for ranges and fixed counts — in your browser, free, no upload. Chapter-shaped documents go to split a PDF by bookmarks, bundled statements to split a PDF by text. Grab all the parts as one ZIP archive, and clean duplex-scan blanks first with auto-remove blank pages.

Related reading

• How to Merge PDF Files — the mirror-image job: many files into one
• Merge and Split PDFs — the family hub: every tool and guide in the family
• File Limits — the sizing rules behind the browser path and the cloud alternative