Split a PDF into smaller PDFs using one of three modes: by page range, into individual pages, or into fixed-size chunks. Everything happens in your browser — no server upload, no file retention, no telemetry. Your PDF never leaves your device.
Three modes
By range(s) — extract specific pages or groups of pages. Input is a comma-separated list like 1-3, 7, 11-15, producing one output PDF per listed range. Useful for pulling a specific chapter or section out of a long document.
Every page — one output PDF per source page. Output N is named page-N.pdf. Useful when downstream processing wants single-page PDFs or when you want to pick out specific pages from a long document afterwards.
Every N pages — fixed-size chunks. A 100-page document split by 25 gives 4 output PDFs. If the source doesn’t divide evenly, the last chunk is smaller.
Example: extracting a chapter
You have a 300-page textbook PDF and only need chapter 4 (pages 87–112). Select the file, choose “By range(s)”, enter 87-112, click Split. You get one output PDF named pages-87-112.pdf with exactly the chapter content. Source is unchanged.
Example: splitting a meeting packet
Your company sent a 20-page meeting packet and you want to email each section separately. Use “Every page” mode to get 20 single-page PDFs, then attach the ones each recipient needs. Total time: under a second.
Example: chunking a large archive
You’ve scanned a 200-page historical document into one giant PDF but need to upload it to a system with a 10 MB file limit. Use “Every N pages” mode with N=50 to split into 4 manageable chunks, then upload them separately.
Upload limits like “10 MB” are almost always decimal (10,000,000 bytes) — match that with what macOS Finder shows. Windows reports the same files using a different unit base, so a chunk that fits the limit on macOS may also report as smaller on Windows. The decimal vs binary byte units explainer covers the math.
Why browser-only
PDFs often contain sensitive information — contracts, medical records, ID scans, financial statements. Uploading them to a random online service is a privacy risk. This tool uses pdf-lib running entirely in your browser: the source PDF is read locally, processed in memory, and each output appears as a direct download. There’s no upload, no queue, no logged retention.
The pdf-lib library (about 400 KB) is loaded lazily on first Split click — it’s not in the initial page bundle, so the page loads quickly even for readers who don’t intend to split anything. After the first load, subsequent splits are essentially instant.
What this tool does not do
It doesn’t re-encrypt or redact the output. If your source has passwords, metadata, or visible content you want gone, handle those with dedicated tools.
It doesn’t bundle outputs into a zip. Each output is a separate download. For many outputs, you could select all from the Downloads folder afterwards — or use a desktop tool if zipping is essential.
It doesn’t rename source pages. The outputs are labelled by page number from the source. For custom naming, rename after downloading.
It doesn’t split by bookmarks or table-of-contents. Splitting by structural outline would require parsing the PDF’s outline tree and is a separate, more complex feature. This tool is page-number-based.
It doesn’t reorder pages. If you need to rearrange pages (e.g., extract and rearrange), extract the ranges first, then merge with the PDF merger. To simply remove pages without extracting, the PDF page deleter is the one-step tool.