About image compression
Image compression reduces a file's size — useful for faster website loading, staying under an upload size limit, or saving storage space — typically by either re-encoding with more efficient lossy compression or reducing quality slightly in ways that are hard to notice visually. This tool compresses multiple uploaded images at once, letting you keep the original format or convert to JPEG/PNG/WebP, and shows exactly how much space you saved.
How the compression works
Each image is decoded onto an HTML canvas and re-encoded using your browser's built-in image encoder at the quality level you set with the slider (0.1 to 1.0 for JPEG and WebP) — lower quality values apply more aggressive lossy compression, discarding more visual detail in exchange for a smaller file. Since PNG re-encoding is always lossless, choosing PNG as your output format won't reduce file size nearly as much as JPEG or WebP for photographic images, though it will still losslessly re-pack the pixel data.
Frequently asked questions
- What quality setting should I use?
- It depends on your priority — values around 0.7-0.85 typically offer a good balance where compression artifacts are hard to notice while achieving meaningful size savings, while lower values (0.4-0.6) save more space at the cost of visible quality loss, useful when file size matters more than perfect visual fidelity.
- Why doesn't compressing a PNG save much space?
- PNG uses lossless compression, so re-encoding it (without converting to a lossy format) can only optimize the existing compression slightly — it can't achieve the large size reductions possible with JPEG or WebP's lossy compression, which is why converting a photographic PNG to JPEG or WebP often gives much better compression results.
- Can I compress an already-compressed JPEG further?
- Yes, technically, by lowering the quality setting further, but each round of lossy JPEG compression discards more image detail — repeatedly compressing the same image can lead to visible quality degradation (generational loss), so it's best to compress from as close to the original, uncompressed source as possible.