PDF Rotate Pages

Rotate individual PDF pages in 90° steps, mark pages for deletion and download a rebuilt file — with pdf-lib in your browser, never an upload.

All processing happens locally in your browser. Your PDF never leaves your device.

Drop a PDF here or click to browse — one PDF at a time

How It Works

Every page of your PDF appears as a small box showing its number and current rotation. Turning a page does not redraw anything: pdf-lib writes the page's rotation entry, and the download step copies each page object verbatim into a new document with that flag applied. Deleted pages are simply left out of the copy.

PDF stores rotation as 90° steps only
A PDF page carries a /Rotate value that is legal only as 0°, 90°, 180° or 270°. Arbitrary angles like 45° are not a rotation — a viewer would have to re-render the page content, which risks quality, so the format forbids it and this tool sticks to quarter turns. Repeated +90° presses cycle through the four legal values.
Deletion rebuilds the file, at original quality
Deleting a page marks it in the editor; on download, pdf-lib creates a brand-new document and copies over only the pages you kept, in order. The surviving page objects — embedded fonts, images, vector paths — are moved unchanged, so nothing is re-encoded and the rest of the document is byte-for-byte as sharp as before.
No thumbnails, no rendering step
The grid shows numbered boxes with rotation labels instead of rendering page previews. That keeps even a 300-page document responsive, because no page is ever drawn to a canvas — the tool only reads the page count and edits rotation flags.

Frequently Asked Questions

Can I rotate just one page?

Yes. Every page has its own +90° button, so you can turn a single upside-down scan without touching the rest of the document. The quick Rotate all left / Rotate all right buttons are there for the opposite case, when every page needs the same turn.

Will rotating reduce the quality of my PDF?

No. Rotation is stored as a flag on the page, and deletion rebuilds the file by copying the surviving page objects verbatim. No page is rendered to an image and nothing is re-compressed, so text, images and vector graphics keep their original bytes.

Why not just let my scanning app auto-rotate the pages?

Auto-rotate guesses orientation from page content and often re-rasterizes the file, which can soften text. It also gets mixed documents wrong — a page that is sideways on purpose stays wrong. Here you set each page's rotation explicitly, and the file is only touched where you clicked.

Why can't I rotate to an arbitrary angle like 45°?

The PDF format only stores page rotation as a multiple of 90 degrees (0°, 90°, 180°, 270°) in its /Rotate entry. That is a spec limit, not a tool shortcut — viewers and printers rely on it, so a 45° turn would mean redrawing the page content, which would degrade quality. This tool therefore only offers quarter turns.

Is anything uploaded to a server?

No. The PDF is parsed, rotated, pruned and saved entirely by pdf-lib running in your browser. There is no upload step, no account and no network request that contains your file.