Compress Images Online — Free Image Compressor

Optimizing your images reduces file sizes by up to 80% without sacrificing quality, leading to faster page loads. This is essential for keeping user attention and ensuring a great user experience. Smaller images also minimize bandwidth usage and improve your website's search engine ranking.

LeanImg uses smart lossy and lossless compression to reduce the file size of your JPG, PNG, WebP, and AVIF images. By selectively optimizing image data, fewer bytes are required to store the image. The effect is nearly invisible but makes a very large difference in file size.

LeanImg supports JPG, PNG, WebP, AVIF, and GIF for compression. You can also convert between these formats. All processing happens in your browser — images are never uploaded to a server. JPG and WebP are ideal for photos, PNG for graphics with transparency, and AVIF for next-gen compression efficiency.

Typical compression reduces file size by 40–80%, depending on the image content and format. Photos with smooth gradients compress very efficiently. You can adjust the quality slider to compress an image to a specific size — for example, under 100 KB for web use — and find the perfect balance between file size and visual quality.

Yes. Select the lossless PNG output format for pixel-perfect compression with zero quality loss. Lossless compression typically reduces file size by 10–40%. For even smaller files, lossy formats like JPG or WebP at quality 75–85 produce results that are virtually indistinguishable from the original.

Yes. LeanImg works as a bulk image compressor — select up to 10 images at a time and compress them all in one go. Once compressed, download them individually or all at once as a ZIP file. Batch compression runs entirely in your browser with no upload required.

Less than you would guess, and that's deliberate. For a JPEG the number is not an absolute target. The compressor reads the quantization tables out of your file, estimates the quality it was already saved at, then applies your setting as a percentage of that. A photo saved at 60 with the slider on 80 comes out at 48, so re-compressing an already-compressed file doesn't quietly re-inflate it. For PNG and WebP input there's nothing to read back, so the number is used as-is.

Yes, and the reason is structural. Compression runs on your own device using WebAssembly, so processing your hundredth image costs us exactly what your first one did, which is nothing. There are no credits to buy, no watermark, no resolution cap and no account. The only limits are practical ones your browser imposes: 10 files and 50 MB in a single batch.

Yes, all of it. Your image is decoded to raw pixels and re-encoded from those, so EXIF, GPS coordinates, the camera and lens fields, any embedded color profile and the thumbnail are all gone from the output. Orientation is the exception, because the rotation is applied to the pixels before we see them. If you need the metadata kept, compress a copy and archive the original.

Usually because it was already optimized. PNG is lossless, so there's a floor below which the file cannot go without discarding color, and a PNG exported by a tool that already ran an optimizer is close to that floor. Our default route for PNG input is a palette quantizer, which does discard color to get the size down. If you picked the lossless PNG option in the dropdown instead, all that runs is oxipng at optimization level 3, and on an already-tight file it has almost nothing left to find.

Better prediction, mostly. WebP's lossy mode is the VP8 keyframe coder, which predicts each block from its neighbors before transforming the difference, so there's less left to store. Google measures around 30% off JPEG at matching quality and that holds up on photographs. It doesn't hold up on everything: WebP stores color at quarter resolution with no option to turn that off, so saturated reds and colored text on a contrasting background can look worse than the JPEG did.

Not directly, though you can get there in one or two passes. Switch to Manual, watch the size on the result row, and move the slider down until you land under your target. As a starting point, one 538.6 KB photo came out at 347.8 KB on quality 90, 212.4 KB on 60 and 131 KB on 30. If you're still over budget after that, resize the image first: halving the dimensions removes three quarters of the pixels and does more for the file size than any quality setting.

Barely, and it costs you quality. JPEG quantizes on a fixed 8x8 grid, so re-saving an untouched image at the same setting mostly converges and stops degrading. Crop it by an amount that isn't a multiple of 8 first, though, and the grid shifts, which forces genuinely new loss on every pass. Compress once, from the best original you have.

For a website you control, yes. AVIF gives the best compression available today, better than 50% off JPEG at comparable quality by Google's measurement, and it carries an alpha channel that JPEG has no way to store. Two costs. Encoding runs the AV1 coder in your browser, so it is the slowest output here by a wide margin, and anything that has to open in older software still wants a JPEG.