Compress a PDF, free
A smaller PDF for email or an upload — by re-encoding the images, not by turning the pages into pictures: the text stays searchable and selectable. Everything happens in your browser.
Drop your PDF here, or choose a file
The document stays in your browser — we never upload it.
You can drop several PDFs at once — up to ten. The finished ones come back as a single ZIP.
What actually makes a PDF smaller?
Almost always the images. In a text PDF the letters are vectors and a full page of text is a few kilobytes; a page photographed with a phone or run through a scanner is a multi-megabyte image the PDF carries verbatim. That is why a ten-page scanned contract can be 20 MB while the same text typed out would be 60 KB.
So the tool does two things. First it re-saves the file: objects are packed into compressed streams, anything nothing points at any more is dropped — every superseded revision of a form that was saved five times — and, if you ask, the metadata goes with it. That step does not change a single pixel. Then it takes the embedded images: any that are larger than the page they sit on needs at the target resolution are scaled down and re-encoded as JPEG.
What it never does is turn the pages into pictures. That is the usual trick behind the big percentages — afterwards the text is no longer searchable, no longer selectable, no longer usable with a screen reader, and a print-ready file becomes a pixel mush. If that is acceptable, a JPEG will do it more cheaply. Here text stays text.
Which setting should you pick?
"Re-save only" leaves the images alone: pick it when the file is going to a printer or somewhere official, or when the sharpness of small type matters. "Light" scales images to 150 DPI — the bottom of printable quality. "Medium" works at 96 DPI: for screens and email that is the right compromise, and it is usually where the biggest saving comes from. "Strong" uses 72 DPI, which is screen resolution: a small file, with visible softening on the images.
DPI is always relative to the page rather than to the image: an A4 sheet at 96 DPI is covered by an image roughly 1120 pixels wide. Anything already smaller than that is not enlarged — a compressor never hands back a bigger image than it was given. Masks and one-bit black-and-white scans are left alone too: JPEG would make those larger, not smaller.
Frequently asked questions
Why did my file not get smaller?
Because there is nothing in it to compress. In a purely textual PDF — an invoice, a contract, a thesis — the space is taken by the fonts and the text, and we do not touch those. What you get then is a few percent from the metadata and the dead objects, or nothing at all. If the result would be larger than the original, we hand back the original and say so.
Does the text stay searchable?
Yes. Text, fonts, bookmarks, links and form fields are carried over untouched — the pages are never redrawn. A scanned document of course has no text in it to begin with, only pictures; there it stays a picture after compression too.
What does removing the metadata delete?
The document information (author, title, producing application, creation and modification times), the XMP packet, the private leftovers editing suites store, the page thumbnails and any files attached to the PDF. Everything that belongs to the content — bookmarks, links, forms, accessibility tags — stays. If you need the attachments, clear the checkbox.
Does it work with a password-protected PDF?
No. An encrypted document cannot be rewritten safely. If you know the password, the PDF password remover can take it off — in your browser too — and the result can then be compressed.
How long does it take?
"Re-save only" is instant. The image settings re-encode every picture separately, so the time depends on how many there are — a 12-page scanned document takes a few seconds on a desktop machine. You can cancel while it runs, and your browser stays usable throughout.
Only need a few pages out of the big file? Use the PDF splitter.