Compress PDF

In a scanned or photographed PDF, nearly all the weight is in the images inside it. This tool finds them one by one, re-encodes them smaller and puts them back, without touching anything else in the document.

That is what separates it from most online compressors, which flatten every page into a photograph: here the text is still text you can select and search, the links still work and the structure survives. On a typical scan the reduction runs from 60 to 90 %.

Images that cannot be re-encoded safely — transparency masks, JPEG 2000, CMYK, one-bit bitmaps — are left exactly as they were. A silently mangled image would be far worse than a file that did not shrink as much.

How to use it

1. Drop in the PDF, or several.
2. Pick a quality, or simply say how much it may weigh.
3. Download the result and check it before you throw the original away.

Frequently asked questions

Is the selectable text lost?

No. Only the embedded images are replaced; the text layer, bookmarks and links are untouched. If the PDF is a scan with no OCR, it never had any.

When should I use the rasterising mode?

When weight matters more than everything else. It redraws each page as a photograph: the drop is dramatic, but the text stops being selectable and searchable, and there is no way back.

My PDF barely shrinks — why?

Because the weight is not in images. A text-and-vector document from a word processor is already compact: there is nothing to squeeze, and the tool says so next to the result.

Is greyscale worth it?

Very much, if the document is scanned text. It drops the colour information, which on a white page with black type is mostly sensor noise, and the result often halves again.

Related tools