Split PDF — Extract Pages, Free & Private
Pull any pages out of a PDF, or split it into several files — every page, every N pages, by ranges or at each bookmark. It all runs in your browser, so nothing is uploaded.
In short: drop a PDF, then either extract the pages you want (1-3, 7, 10-) into one new document, or split it into several files — every page, every N pages, one file per range, or at each bookmark. It runs on your device, so nothing is uploaded and there is no page or size limit, and internal links and form fields come out working, which is not true of most tools that do this.
- Is it free, and is there a limit?
- Free, with no task limit, no page limit and no size limit. Sejda allows 3 tasks an hour on its free tier and Smallpdf 2 a day; there is no server here, so there is nothing to meter.
- Is my PDF uploaded anywhere?
- No. The split happens in your browser, so the document never leaves your device — which matters, because the things people split are contracts, statements and medical records.
- Do links and form fields survive?
- Yes, and this is worth checking elsewhere. Splitting with the usual library call breaks both silently; this page repairs them and says how below.
- Can I extract non-consecutive pages, like 1, 4 and 9?
- Yes. Type them in the pages box —
1, 4, 9or1-3, 7, 10-— and they come out as one PDF in the order you typed. - Can I split into several files at once?
- Yes. Split into files makes one PDF per page, per N pages, per range you list, or per top-level bookmark, and downloads them as one ZIP.
- Does it change the pages themselves?
- No. The pages you keep are copied across exactly as they were — same text, same images, same quality. Nothing is re-rendered.
How to Split a PDF
Drop it onto the tool, or click to browse. Nothing is uploaded.
Extract pages such as 1-3, 7 into one PDF, or split into files every page, every N pages, by ranges or at bookmarks.
The pages are copied on your device, with links and form fields repaired.
Save the new PDF, or a ZIP of several.
Features
Any pages
Type 1-3, 7, 10- and get one PDF of just those.
Several files
Every page, every N pages, or one file per range, as one ZIP.
At bookmarks
A new file at each top-level bookmark, named after it.
Nothing breaks
Internal links and form fields come out working.
Protected PDFs
Restricted and password files open.
Nothing uploaded
It all runs in your browser.
What splitting normally breaks, and what this repairs
Splitting a PDF looks like the simplest operation there is — keep some pages, drop the rest — and the library call that does it quietly breaks two things every time. This page did too, until September 2026, and most tools still do.
Internal links
A link from a contents page to chapter three is stored as a reference to that page object. When pages are copied into a new document, the destination gets deep-copied along with the link — so the link now points at an orphaned duplicate of chapter three that is not in the file, rather than at the chapter three that is. Measured on a real document: one working link in, zero working links out, even when the target page was inside the range that was kept.
This page records where every internal link pointed before copying anything — the references mean nothing afterwards — and re-points them at the pages that really are in the output. A link to a page you left out is deleted rather than left aimed at whatever now happens to sit at that number, because a link that quietly goes to the wrong page is worse than one that does nothing.
Form fields
A fillable box on the page is a widget; the field it belongs to is a separate object the widget points at. Copy the pages and the widgets come along, but nothing registers the fields — so every reader shows the boxes as decoration. Measured: one fillable field in, zero out. This page follows each widget to its field and registers it, so the extracted pages are still a working form.
Bookmarks
These really are lost, and cannot be recovered this way — so the capability table says No rather than something vaguer. If the bookmarks matter, keep the whole document.
Why do it in the browser
Every large PDF site does this on a server, which means your document travels to a machine you do not control so that a few pages can be copied out of it. The limits they impose are the tell: they exist to protect bandwidth and CPU, not to help you.
Checked in September 2026: Sejda’s free tier allows 3 tasks an hour, caps documents at 50 pages and 50 MB. Smallpdf allows 2 tasks a day. iLovePDF allows roughly 1–2 an hour and caps Split at 100 MB. PDF24 is genuinely unlimited and free — and its web version uploads too, though it also ships a desktop app that works locally.
Here the split is a page-copy performed by JavaScript in your own tab. There is nothing to meter, so there are no tasks per hour, no page cap and no size cap; and the documents people bring to a splitter — contracts, bank statements, medical letters, tenancy agreements — never leave the machine they were already on.
Capabilities at a glance
What this tool does and does not do, stated plainly. The “No” rows are the point, not an omission.
| Capability | Supported | Detail |
|---|---|---|
| What it does | ||
| Extract one continuous page range | Yes | First and last page; the rest is left behind. |
| Non-consecutive pages (1, 4, 9) | Yes | Type 1-3, 7, 10-; one PDF, in the order typed. |
| Splitting into several files at once | Yes | Every page, every N pages, or one file per range; one ZIP. |
| Split at bookmarks | Yes | One file per top-level bookmark, named after it. |
| Pages copied exactly, not re-rendered | Yes | Same text, images and quality as the original. |
| Internal links inside the range keep working | Yes | Repaired explicitly — the usual library call breaks them. |
| Links to pages you left out | No | Deleted rather than left pointing at whatever now sits at that number. |
| Form fields stay fillable | Yes | Also repaired — they come out unusable otherwise. |
| Bookmarks kept inside each file | No | Lost in the split, and they cannot be recovered this way. |
| Reordering pages | No | Organize PDF reorders, deletes and keeps any set. |
| Input | ||
| Any PDF | Yes | Including scans and forms. |
| A password-protected PDF | Yes | Restricted files open automatically; an open password is asked for and stays on your device. |
| Page limit | No | Sejda caps the free tier at 50 pages; there is no cap here. |
| File size limit | No | There is no server to protect. iLovePDF caps Split at 100 MB free. |
| Task limit | No | Smallpdf allows 2 free tasks a day, Sejda 3 an hour. |
| Output | ||
| A normal PDF of just those pages | Yes | Opens anywhere. |
| Several files as one ZIP | Yes | Named after the pages, or after the bookmark. |
| The original left untouched | Yes | The result is a new download. |
| Watermark | No | Nothing is added to the pages. |
| How it runs | ||
| Nothing is uploaded | Yes | The document never leaves your device. |
| Signup | No | There is no account. |
| Offline use | No | The page and pdf-lib load from the network first. |
Extract pages, or split into files
Extract pages makes one new PDF. Give a first and last page, or type any mix in the pages box — 1-3, 7, 10-, where 10- runs to the end. Pages come out in the order you typed them, and a page typed twice appears once.
Split into files makes several PDFs in one ZIP:
- Every page — one PDF per page.
- Every N pages — 30 pages at 10 gives three files; the last one takes whatever is left.
- One file per range —
1-3, 4-9, 10-gives three files, for chapters you already know the page numbers of. - At each bookmark — reads the document’s top-level bookmarks and starts a new file at each one, named after it. Pages before the first bookmark become their own file. A PDF without bookmarks is told so rather than split some other way.
The count of files appears beside the setting before you press anything. If you want to reorder pages or delete a few from the middle while keeping the rest, Organize PDF does that with thumbnails, and runs the same repairs.