How our PDF compression works
Most compressors rebuild your document from scratch. That is why text often comes back softer, fonts get substituted, and pages you could once select text on turn into flat images. We take a different route: we rewrite only the images and the fonts, and leave your text and vector artwork untouched, byte for byte. You get a much smaller file that still reads, searches and prints exactly like the original.
Images are resized to the space they actually occupy
This is where most of the size goes, and where most tools get it wrong. A 4000-pixel-wide photo placed two inches wide on the page carries around 2000 dots per inch — many times more detail than any screen or printer can use. We measure how much room each image really takes up on its page and resize it to suit that space. Conventional tools apply one fixed number to every image, which ruins the photo that fills a page while leaving the wasteful small one untouched.
Repeated images are stored only once
A logo that appears on all 40 pages of a report is usually stored 40 separate times. We fingerprint every image, keep a single copy, and point every page at it. In our testing this alone took a 349 KB report down to 16 KB.
Every image gets the encoding that suits it
Photographs are encoded as JPEG. Flat artwork — screenshots, charts, diagrams and logos — is kept completely lossless, because JPEG blurs hard edges on this kind of content and usually produces a larger file, not a smaller one. Black-and-white scans keep their original encoding, which is already more efficient than anything we could replace it with.
Fonts are reduced to the characters you actually used
A PDF embeds an entire typeface even when the document uses forty characters of it. We keep only the characters that appear and discard the rest, while keeping the font embedded so the document still looks identical everywhere. On a tax form containing no images at all, this step alone cut 27% of the file size.
The file structure is cleaned up
We remove objects nothing refers to any more, compress the document's internal structure, and rewrite the file for Fast Web View, so the first page appears while the rest is still downloading.
Choosing a compression level
| Level | Best for |
|---|---|
| Extreme — less quality, high compression | Sending by email or messaging apps, where file size matters more than fine image detail |
| Recommended — good quality, good compression | The default choice for almost everything: images stay sharp on screen with a large size reduction |
| Less compression — high quality, less compression | Files intended for printing or long-term archiving |
The three levels differ in how images are handled only. Text, fonts, links and bookmarks are treated identically at every level.
What stays intact
- Selectable, searchable text — pages are never converted to images
- Embedded fonts — reduced in size, but never dropped or substituted
- Internal links and bookmarks — every destination still resolves
- Form fields and annotations — preserved as they were
- Your original file — if compression cannot make it smaller, we return it unchanged