Shrink scanned documents and image-heavy PDFs 4–10× — each page is re-rendered and re-encoded entirely in your browser.
Scanned documents are the sweet spot: scanner software often saves pages as oversized, lightly-compressed images, and re-encoding them at a sensible resolution typically cuts the file to 10–25% of its original size. Photo-heavy PDFs and exported slide decks also compress well. PDFs that are mostly vector text (like a report saved from Word) are already tiny — re-rendering those would make them bigger, so this tool detects that case and tells you honestly instead of handing you a worse file.
Each page is rendered to a bitmap at your chosen resolution, encoded as an optimized JPEG, and rebuilt into a fresh PDF with the same page dimensions. That fresh file also sheds accumulated baggage — old metadata, embedded fonts, duplicate objects. The trade-off: text on the page becomes part of the image, so it can no longer be selected or searched. For scanned paperwork nothing is lost; for born-digital documents, keep the original alongside the compressed copy you email.
The whole pipeline — rendering, encoding, rebuilding — runs in your browser tab. Your document never leaves your device: nothing to upload, nothing cached on a server, nothing to delete later. Document metadata is stripped in the output as a side effect of the rebuild.
Encrypted PDFs can't be opened directly — remove the password in your PDF reader first, then compress here.
JPEG size depends heavily on page content, so promising an exact size would mean compressing repeatedly and burning time. The three quality levels are calibrated so that Balanced lands scans around 10–25% of the original; if you need smaller, try Strong or grayscale.