Compress Image — Reduce File Size, Keep Quality

For JPEG, going from 100% to 80% quality typically reduces file size by 50–70% with virtually no visible difference on screen or print. For PNG with flat colors, switching to WebP lossless can reduce size by 25–35%. The right quality depends on your use case — web thumbnails can use 60–70% quality; print assets should stay at 85–95%.

JPEG uses lossy compression — pixels are mathematically approximated. Lower quality = smaller file but some detail is lost. PNG uses lossless compression — every pixel is preserved exactly, but file sizes are larger. PNG compression only removes redundancy in the data, not actual image information. For photos, JPEG or WebP is more efficient; for screenshots, icons, or images with transparency, PNG or WebP lossless is appropriate.

Yes, re-compressing JPEG introduces "generation loss" — each compression cycle degrades quality slightly. This tool minimizes damage by using a high-quality decoder (jsquash), but the principle applies. If preserving maximum quality is critical, compress from the original RAW or lossless source, not from an already-compressed JPEG.

PNG lossless compression has physical limits — complex photos with millions of unique colors can only compress so much without losing data. For photos, WebP lossy or JPEG will always be smaller. PNG excels at artificial images: screenshots, diagrams, logos, and UI elements with few unique colors, where compression ratios of 5–10× are achievable.

Standard JPEG compression is inherently lossy. There are lossless JPEG optimization tools (like jpegtran) that strip metadata and optimize Huffman coding without touching pixels — typically saving 2–8%. This tool uses lossy re-encoding at your chosen quality. For lossless JPEG optimization, consider ImageOptim on macOS or Squoosh with its Mozjpeg encoder at 100% quality.