PDF Compress

PDF compressor

PDFs get big when they’re full of scanned images, embedded photos or large fonts. This compressor re-encodes the images inside the PDF (usually the main size culprit) using JPEG or JPEG2000 at a chosen quality level, subsets embedded fonts to include only used glyphs, and removes duplicated resources. Typical savings: 40-80% on image-heavy files, 20-40% on text-heavy ones.

How compression works

  1. 1

    Upload the PDF

    Any document up to common online size limits.

  2. 2

    Pick quality level

    High (light compression, near-original), medium (balanced), low (maximum compression).

  3. 3

    Image re-encoding

    Embedded images are recompressed at the chosen quality.

  4. 4

    Font subsetting

    Embedded fonts keep only the glyphs actually used in the document.

  5. 5

    Resource deduplication

    Repeated images (e.g. a logo on every page) are stored once and referenced.

  6. 6

    Download compressed file

    Compare before/after size in the result view.

Typical compression ratios

PDF type Before After (medium) Savings
Scanned contract (300 dpi) 25 MB 4 MB 84%
Photo-heavy catalogue 40 MB 12 MB 70%
Mixed report (text + charts) 8 MB 3.5 MB 56%
Text-only book 2 MB 1.6 MB 20%
Born-digital invoice 200 KB 180 KB 10%

Born-digital PDFs (exported directly from Word, InDesign, web-to-PDF) are already fairly efficient; scanned PDFs have the most room to shrink.

Quality settings, explained

What gets compressed

What doesn’t compress

Tips for email and upload limits

Target Typical PDF limit
Gmail attachment 25 MB
Outlook attachment 20 MB
WhatsApp file 100 MB
LinkedIn message 20 MB
Most web upload forms 10 MB (varies)

Compressing to 5 MB or less gives comfortable margin for email and most uploads.

When not to compress

Frequently Asked Questions

Text in PDFs is usually stored as actual text (glyphs), not images, so text quality is unaffected by image compression. Only if the PDF is a scan where every page is an image does text readability depend on compression level.

Medium quality is usually the sweet spot — strong size reduction (40-70%) with minimal visible quality loss. Drop to low only if medium doesn’t fit the limit.

Compression can alter pixel-level image data, which matters only in a tiny number of cases (forensic evidence, artwork reproduction). For contracts, reports and general documents, compression doesn’t affect content or signing validity.

Yes, but expect diminishing returns. Double-compressing JPEG images degrades quality further each time. Start from an uncompressed source whenever possible.

Related Tools