PDF Tools
Four PDF jobs, one panel
Combine several PDFs, break one apart, fix scanner rotation, or squeeze a file under an attachment limit.
PDF Tools
Runs in your browser · nothing uploaded01 — How to use PDF Tools
02 — Getting good results
03 — Questions
No. Every file is read and processed entirely inside this browser tab, using pdf-lib on your device — there is no upload step, and no network request happens while a file is merged, split, rotated or compressed. This is a deliberate change from the old version of this tool, which sent documents to a server.
It depends on what the PDF contains. Scanned documents and photo-heavy files usually shrink a lot at Balanced or Smallest, because their size is almost all JPEG images. A text-only PDF exported from a word processor is often already compact, and Compress will say so and hand back the original rather than a file that is bigger. It never returns a larger file than the one you gave it.
Only JPEG images in RGB or greyscale are re-encoded. CMYK images from print workflows, and images stored in other formats, are left exactly as they are: a browser cannot decode CMYK JPEGs with faithful colours, and re-encoding them would visibly shift the colours of your document.
Each PDF can be up to 100 MB. Merge takes up to 50 files at once; Split, Rotate and Compress take one. A merged output has no separate size cap, so check its size in the result bar before sending it — and run it through Compress if it is too big.
04 — More about PDF Tools
How a browser can edit PDFs without a server
A PDF is not a flat stream of pages — it is a graph of objects (page dictionaries, content streams, fonts, images) linked together and indexed at the end of the file. Merging builds a new graph containing every page from every source; splitting builds several, each holding a subset; rotating changes one number on each chosen page. pdf-lib does all of this in JavaScript inside the tab, the same kind of parsing and byte-level work a desktop PDF editor does in a native process.
What Compress actually does
Three things, in order. First it walks the document from its root and deletes every object nothing refers to — editors that save by appending changes leave old versions of pages, fonts and images behind, and those can be a large share of the file. Second it compresses any stream stored with no compression at all. Third, at Balanced and Smallest only, it decodes each JPEG photo, scales it down if it is larger than the level allows, and encodes it again at a lower quality, keeping the new version only when it is genuinely smaller. Objects are then packed into compressed object streams as the file is written.
Why rotation does not lose quality
Every PDF page carries a /Rotate value telling viewers how to orient it. Rotate adds 90, 180 or 270 degrees to that value and changes nothing else: text stays selectable, vector art stays vector, and scanned images are not resampled. The legacy version of this tool redrew each page under a rotation transform instead, which worked but rewrote content that never needed to change.
What password protection blocks
PDF encryption scrambles the file’s content streams and, depending on how it was protected, may also restrict copying or printing. All four operations need to read those streams, which is impossible without decrypting them. This tool intentionally does not attempt to guess or brute-force a password; a protected file fails with a message naming it, and must be unlocked elsewhere first.
61 more tools
where this came from.
Same shell, same shortcuts, same promise: nothing leaves your browser.