Compress a PNG to 100KB

Processed entirely on your device — images are never uploaded

PNG and JPG solve different problems, and that difference becomes very obvious the moment you try to hit a specific file size target. A JPG compressor has a dial to turn — quality — that trades visual fidelity for file size in a smooth, predictable way. A PNG compressor doesn’t have that dial at all, because PNG is a lossless format: it guarantees that every pixel you get back is bit-for-bit identical to the pixel you started with. That guarantee is exactly why PNG is the right format for screenshots, logos, and anything with sharp text or hard edges — but it also means shrinking a PNG to a target size like 100KB requires a genuinely different approach than shrinking a JPG does.

What “lossless” actually limits you to

Since a lossless format can’t approximate or simplify pixel data the way JPG does, the only two levers available for making a PNG smaller are:

  1. Smarter lossless re-encoding. PNG’s compression (built on the same family of algorithms as zip/deflate) can often be tuned more aggressively than whatever default settings produced the original file — many PNGs, especially ones exported quickly from screenshot tools or design software, are not compressed as tightly as the format actually allows. Re-running a more thorough compression pass over the exact same pixel data can shrink the file meaningfully without changing a single pixel.
  2. Reducing pixel dimensions. If lossless re-encoding alone can’t reach the target, the only remaining option is to have fewer pixels to encode in the first place — i.e., making the image physically smaller. This genuinely changes the image (it’ll display at a smaller size, or need to be scaled up by whatever displays it), unlike the first option.

This tool tries the first approach exhaustively before ever touching the second, since reducing dimensions is the more disruptive change. For most typical screenshots and graphics, the lossless pass alone gets surprisingly far.

Why a PNG can be much bigger than a “similar” JPG

It’s common to compare a PNG and a JPG of what looks like the same content and be surprised the PNG is many times larger. The difference comes down to what each format is actually compressing well. JPG’s lossy compression is built around smoothly varying photographic detail — it can approximate a sky’s gradient or skin tones with very few bits while still looking convincing. PNG’s lossless compression is built around exact repetition — large flat areas of identical color compress extremely well, but anything with continuous tonal variation (a photo, a gradient, a busy textured background) gives it much less to work with, since it has to represent every subtle variation exactly rather than approximating it away.

This is why a flat-color app icon or a simple diagram can be a tiny PNG, while a screenshot containing an embedded photograph or a smooth gradient background can balloon to several megabytes even at modest pixel dimensions — and why, if your “PNG” is actually a photo someone saved in the wrong format, converting it to JPG first is almost always the better path to a small file size than trying to compress the PNG itself.

When you should actually use the PNG-to-JPG converter instead

If transparency and pixel-perfect sharp edges don’t matter for your specific image — most photographs don’t need either — converting to JPG before compressing will almost always get you a smaller file at a given quality level than staying in PNG ever can, simply because JPG’s lossy compression is so much more efficient at handling photographic detail. Keep this PNG-specific compressor for the cases PNG actually exists for: screenshots with text, logos and icons with transparent backgrounds, and diagrams or UI mockups where sharp, exact edges matter more than absolute minimum file size.

How it works

  1. Add your PNG

    Works well for screenshots, app icons, logos with transparency, and diagrams — the kind of flat-color, sharp-edged graphics PNG is actually designed for.

  2. The tool optimizes losslessly, then resizes if needed

    Unlike a JPG target, there's no quality dial to turn down — PNG is lossless. The tool first runs a lossless re-compression pass, and if that alone can't reach 100KB, it reduces pixel dimensions and tries again.

  3. Download — transparency stays intact

    The output is still a PNG, so any transparent areas in your original (a logo on a transparent background, for instance) are preserved rather than flattened onto a solid color.

Frequently asked questions

Why does compressing a PNG work differently from compressing a JPG?

JPG is a lossy format — it can throw away visual detail in exchange for a much smaller file, controlled by a quality setting. PNG is lossless by design: every pixel comes back out exactly as it went in, which is exactly why PNG is the right choice for screenshots and graphics with sharp text or hard edges. The tradeoff is that PNG has no quality dial to turn down, so the only ways to make one significantly smaller are smarter (but still lossless) compression of the existing pixel data, or reducing how many pixels there are in the first place.

Will this lose my PNG's transparency?

No — the output stays in PNG format specifically so that any alpha transparency in the original is preserved. If you converted to JPG instead, transparency would be lost entirely (JPG has no concept of it), which is the main reason to use this tool rather than a generic compressor when transparency matters.

My PNG is a photo, not a graphic — should I still use this?

Probably not, if file size is your priority. PNG is a poor fit for photographic content — its lossless compression can't shrink smooth gradients and fine detail nearly as efficiently as JPG's lossy compression can, so a photo saved as PNG is often 5-10x larger than the same photo saved as a reasonably-compressed JPG. If you don't specifically need transparency or lossless edges, converting the photo to JPG first will get you a much smaller file at a target size like 100KB than compressing the PNG itself ever could.

Why is my screenshot PNG so big in the first place?

Screenshots, especially of detailed UIs, photos embedded in a page, or anything with a lot of color variation, give PNG's lossless compression relatively little repetition to exploit, which is what that compression actually relies on to shrink files. A screenshot of a mostly-text interface compresses far better than a screenshot containing an embedded photo or a complex gradient background, even at the same pixel dimensions.

If lossless compression alone isn't enough, what actually happens to my image?

The tool reduces the image's pixel dimensions and re-compresses, repeating with progressively smaller dimensions until the result fits under 100KB or it hits a sensible minimum size. You'll get a smaller image, not a lower-fidelity one at the same size — every pixel that remains is still exactly accurate, there's just fewer of them.

Related tools