Compress PNG Online

How to Compress a PNG Image

How PNG Compression Works

PNG uses the DEFLATE compression algorithm — the same algorithm behind ZIP files. It’s a lossless method, meaning every single pixel is preserved exactly as in the original. No data is ever discarded.

Before DEFLATE runs, PNG applies a filter strategy to each row of pixels. The filter predicts pixel values based on neighboring pixels and stores only the difference. Five filter types are available:

• None — stores raw pixel values (best for random noise)
• Sub — stores difference from the pixel to the left
• Up — stores difference from the pixel above
• Average — uses the average of left and above pixels
• Paeth — uses a linear function of left, above, and upper-left pixels

Our tool tests multiple filter strategies and DEFLATE parameters to find the combination that produces the smallest file. This is why re-compressing a PNG from another tool can sometimes yield further savings — the original may not have used optimal settings.

PNG vs JPG: When to Use Each

Use PNG when you need transparency, pixel-perfect rendering, or sharp text. Use JPG when file size matters more than pixel accuracy — especially for photographs.

Tips for Smaller PNG Files

• Reduce the color palette — if your image uses fewer than 256 colors (icons, simple graphics), an 8-bit PNG with an indexed palette can be 3–5× smaller than a 24-bit PNG.
• Remove metadata — PNG files often contain EXIF data, ICC color profiles, and text chunks. Stripping these can save 10–50 KB per image.
• Use 8-bit instead of 24-bit — for images with limited colors (diagrams, UI screenshots with flat design), 8-bit color depth is visually identical but dramatically smaller.
• Simplify gradients — smooth gradients create many unique pixel values, making compression less effective. If possible, use solid colors or stepped gradients.
• Consider WebP — WebP supports both lossless and lossy compression with transparency. Lossless WebP is typically 26% smaller than PNG for the same image.