Why are GIFs so big, and how do we shrink them?
GIF uses LZW compression on a 256-color indexed palette. It is a 1989 format and was not designed for the kind of full-color screen recordings people throw at it today. Most bloated GIFs share three problems: a global palette applied to every frame, no frame disposal optimization, and no removal of redundant or near-duplicate frames. Gifsicle solves all three.
Pixshrink runs gifsicle in WebAssembly. It analyzes each frame, picks a per-frame palette when needed, optimizes transparency, and merges or drops frames that contribute nothing. The visible animation is identical; the file is dramatically smaller.
How much smaller should I expect?
On a 480×270 screen-recorded GIF, 8 seconds at 15 fps, gifsicle typically reduces the file by 30–60% with no visible change. The first encode is lossless — the exact same pixel output, just stored more efficiently. If you accept lossy color reduction (dropping from 256 to 128 or 64 unique colors per frame), you can push much further; the encoder picks the right palette automatically using a NeuQuant-style algorithm.
If even optimized GIF is too large for your use case, the right next step is to convert to animated WebP — typically 4–8× smaller than the equivalent GIF at the same visual quality — and check that your destination supports it.
Can I drop frames or change the loop?
Yes. The options panel lets you set a target frame rate (skip every Nth frame), force a global palette, and tweak how aggressive the optimizer is. You can also reverse, scale, or crop the animation without re-encoding through a separate tool.
When to switch formats
GIF is the most compatible animation format in existence. If your audience includes users on platforms that do not decode video (rare, but real), GIF is still the right answer. For the web, social media, and most apps, animated WebP or short AVIF sequences beat GIF on size, color depth, and battery cost. Pixshrink outputs both from the same source.