How to Compress a PDF Without Losing Too Much Quality
Updated · 6 min read · MetaSol Technologies
A 40 MB PDF is almost never 40 MB of text. A page of plain text takes a few kilobytes. The weight comes from images: scans, photos and screenshots stored at a higher resolution than any screen or printer needs.
What compression actually does
Compress PDF looks at every image in the document and:
- Reduces its resolution if it is larger than needed for the chosen level.
- Re-encodes it as JPEG at a controlled quality.
- Keeps the original image if the new version would not be at least 10% smaller.
Text, vector drawings, fonts and links are not touched, so they stay perfectly sharp at any zoom. If the whole document cannot be made smaller, you get your original file back, never a bigger one.
Choosing a level
| Level | Typical use | What you might notice |
|---|---|---|
| Light | Brochures, portfolios, anything printed | Usually nothing |
| Recommended | Email, sharing, on-screen reading | Slight softening of photos at high zoom |
| Strong | Upload portals with 2–5 MB limits | Visible softening in photos; text unaffected |
Start with Recommended. If the file is still too large, try Strong.
Scanned documents
Scans are pages of images, so they compress very well. For the best results at the source:
- Scan text documents at 200–300 DPI, not 600.
- Use grayscale for black-and-white paperwork. It is often a third the size of colour.
When compression won't help much
- Text-only PDFs (exported from Word, for example) are already compact.
- CMYK print files keep their images untouched to protect print colours.
For these, the lossless Optimise PDF can still remove unused data. Splitting the document with Split PDF is the other reliable way to meet a size limit.
Privacy
Compression runs on our server because it needs image codecs that browsers don't provide to web pages. Files are processed in a private temporary folder and deleted the moment the result is returned.