Split a PDF

PDFSplit a PDF into separate files, extract a page range, or save every page as its own PDF, in your browser. Free, no upload, no watermark, no page limits.

Your PDF is split on this device and never leaves it. Nothing you open is ever uploaded. There is no server to upload it to.

What this does

It takes one PDF and gives you back less of it, or the same pages in better-sized pieces: a single new PDF containing just the pages you choose, a separate file for each range you mark out, or every page as its own PDF. The result downloads as one PDF when there is one output file and as a zip when there are several, and nothing about your original changes.

As with every tool on this site, the work happens in your browser. Sites that split PDFs on their servers require the one thing a bank statement, a contract or a medical letter should not do: travel. Here there is no server to receive the file, no queue to sit in and no retention policy to trust, because the pages are copied apart on your own processor and handed straight back.

How to use it

  1. Drop your PDF onto the tool, or press “Choose a PDF”. The file’s page count and size appear once it has been read.
  2. Pick the shape of the output: one PDF with chosen pages, a separate PDF per range, or every page as its own file.
  3. Type the pages, unless you picked every-page, in which case there is nothing to type. The line under the box shows live exactly what you will get, named file by file, or tells you precisely what it could not read, down to “this document has 12 pages, so there is no page 14”.
  4. Press “Split PDF”, then download. One output file arrives as a PDF; several arrive as a zip, with the contents listed so you know what is in it before you open it.

The range language, in full

The Pages box accepts pages and ranges separated by commas: 7 is a page, 2-5 is inclusive, 12- runs to the end of the document, -3 means from the beginning. Spaces do not matter.

Two behaviours are deliberate rather than accidents. Order is preserved, so 5, 1-2 in extract mode produces a PDF whose first page is page five, which quietly makes the extract mode a page re-arranger. And repeats are honoured, so 1, 1 really does give you the cover twice; the tool assumes you mean what you typed and shows you what it understood before anything is produced.

Output names follow the input, so a folder of results stays recognisable: report.pdf yields report-page-7.pdf, report-pages-2-5.pdf, report-selected-pages.pdf when several ranges go into one file, and report-split.zip when they do not.

What survives the split, and what cannot

Pages come through exactly. Each page is copied as an object with everything it references: content, fonts, images, colour profiles, dimensions. Nothing is rasterised or re-encoded, a searchable page stays searchable, and an A4 page is still precisely A4. File sizes are roughly proportional to what each piece contains, with shared resources copied into every piece that needs them.

Document-level features cannot. Bookmarks and the machinery that makes form fields fillable belong to the document’s catalog, not to its pages, and a piece of a document does not inherit them: outputs carry no outline, and forms keep their appearance, including anything already typed, but stop being interactive. Most splitting tools quietly do the same; the difference is that this page says so before you find out.

Encrypted files are refused at the door. A surprising number of PDFs carry restrictions you have never noticed, because viewers open them without a password. Copying pages out of a file this tool cannot decrypt would produce broken output, so it declines and says why.

When less is the right amount

The classic case is sending one thing without sending everything: page three of the statement for the landlord, the signed page of the contract, one chapter of the manual for the person who asked. Scanned bundles want the opposite treatment, broken back into the documents they actually are, one range per document. And every-page mode suits the paperwork that gets filed page by page: invoices scanned in a batch, certificates collected into one PDF that should never have been one PDF.

Limitations

One document at a time. This tool splits a single PDF. Combining several into one is the PDF merger, and the two undo each other cleanly.

The interactive layer is not carried. As above: no bookmarks or fillable fields in the output. The pages themselves, and how they look, are untouched.

Protected files need their password dealt with first. The tool refuses encrypted PDFs rather than producing a broken split, so remove the protection in software that has the password, then split.

Memory is the ceiling. The document and its pieces are held in your device’s memory while you work; that is the price of the no-upload architecture. Ordinary documents are invisible; a two-thousand-page archive split page by page may not be, and batches are the answer.

Frequently asked questions

How do I split a PDF into multiple files?

Drop the PDF onto the tool, choose "A separate PDF for each range", and type where the cuts go, such as 1-4, 5-9, 10- for three files with the last one running to the end. Each range becomes its own PDF, and they arrive together in one zip. If you want every page as its own file, that is the third option and needs no typing at all.

How do I extract just one page from a PDF?

Drop the file in, leave "One PDF with the pages you choose" selected, and type the page number, say 7. The download is a one-page PDF named after the original, report-page-7.pdf for report.pdf, and the original file on your machine is untouched.

Is it safe to split a PDF online?

That depends on where the splitting happens. Most PDF sites work by uploading your document to their server, and the documents people split, statements, contracts, medical records, are exactly the ones that should not travel. This tool runs in your browser: there is no server that could receive your PDF, and you can watch the network panel while you split to see that nothing carrying it leaves.

Will the split pages lose quality?

No. Pages are copied as objects, not re-rendered: text stays selectable, images keep their original compression, fonts travel with the pages that use them, and every page keeps its exact size. There is no quality setting because no quality decision is being made.

Can I split a password-protected PDF?

No, and the refusal is deliberate rather than a missing feature. Even a PDF that opens without asking you anything can carry owner-password encryption, and copying pages out of a file this tool cannot decrypt would produce garbage. Remove the protection first in a viewer you have the password for, then split the unprotected copy.

What can I type in the Pages box?

Pages and ranges, separated by commas. 7 is a page, 2-5 is pages two to five, 12- runs from page twelve to the end, and -3 means from the start. Order matters and repeats are allowed, so 3, 1, 2 extracts three pages in that order. If something cannot be read, the line under the box says exactly what and why, including the document's real page count.