How PDF compression actually works
A PDF is a container. Its size is the sum of its parts: embedded fonts, vector drawing instructions, the text itself, and — usually dominating everything else — raster images. A five-page scanned contract is really five photographs wrapped in PDF structure, which is why it can weigh 20 MB while a fifty-page text document sits under 200 KB.
That distinction decides what any compressor can do for you. This tool re-encodes the embedded JPEG images: each one is decoded, scaled down if it is wider than the chosen limit, and written back at a lower quality setting. It then rewrites the file structure using object streams and, if you ask, discards the document metadata.
What to expect from each preset
- High quality keeps images up to 2200px wide at 82% JPEG quality. Use it for documents that will be printed or read closely.
- Balanced caps images at 1600px and 65% quality. This is the right default for email attachments and uploads to portals with a size limit.
- Smallest file caps at 1000px and 45% quality. Text stays crisp because text is not an image, but photographs will show visible JPEG artefacts.
Honest limits
If your PDF is text and vector graphics, there is very little to remove and the saving will be a few percent at most. Any tool that claims a large reduction on such a file is either re-saving it with no real change or rasterising your text into images, which destroys selectable text and accessibility. This tool does neither — it reports how many images it found and how many it actually re-encoded, so you can see exactly where the saving came from.
Images the browser cannot decode, such as CMYK JPEGs used in print workflows, are left untouched rather than mangled. Encrypted PDFs must be unlocked before compression.
Everything happens inside your browser tab. The file is read into memory, rewritten, and offered back as a download — it is never transmitted anywhere, which matters when the document is a signed contract, a medical record, or an identity document.