How to reduce a PDF's file size — without wrecking it

A PDF that won't attach to an email or clears a strict upload limit is almost never bloated because of the text. Text is tiny — a 40-page report's worth of words is a few kilobytes. The size comes from three places: embedded images kept at their original camera or scan resolution (a phone photo dropped into a page can be 4-8MB on its own), fonts embedded in full even when only a handful of characters are used, and, for scanned documents, each page effectively being a full-resolution photograph with a thin text layer stapled on top.

That's what a PDF compressor is actually doing when you run one: finding the embedded images and re-encoding them at a lower resolution and JPEG quality, and stripping font subsets and duplicate objects the file doesn't need. It isn't a smarter version of zip compression — there's no clever trick that shrinks a file without touching what's in it. Every meaningful size reduction comes from throwing away image detail you probably won't notice, in exchange for detail you definitely will if you push it too far.

That's also why compression has a ceiling that depends entirely on what's inside the file. A text-only contract with no images might drop by 5-10% and that's genuinely the floor — the images-are-the-weight rule means there's nothing left to compress. A PDF built from twenty full-resolution photos can often shrink by 80-90% with no visible difference at normal reading size, because those photos were carrying far more resolution than a screen or printed page needs anyway. If a compressor promises the same aggressive result on both, one of those two claims is wrong.

The practical approach: use a quality slider rather than a single "compress" button when one's available, and check the result at the resolution you'll actually view it at — full-screen on a laptop, not zoomed to 400% checking for artifacts nobody will ever look for. Our Compress PDF tool exposes that slider for exactly this reason, rather than picking one aggressiveness level and hoping it's right for your file.

The other reason this is worth doing in a browser rather than one of the upload-based compressors that dominate search results: the moment you're compressing a PDF to get it under a size limit is very often the moment you're about to send it somewhere — attach it to an email, upload it to a portal, share it with someone. That's not a neutral file. Running it through a third-party server first, right before it goes somewhere that actually matters, is the opposite of reducing its exposure.