About
Fast, exact, and private by construction.
The problem with “compressed”
Almost every image tool optimises for a percentage. Almost every form enforces a byte limit. Those two things are not the same, and the gap is where people lose an afternoon: submit, reject, lower the slider, submit again. A tool that takes the limit as an input removes that loop entirely.
How it is built
- Static HTML generated at build time — no server-side rendering, no hydration.
- No front-end framework. The page is HTML and CSS; interaction is about 9 KB of plain JavaScript.
- Image work runs in Web Workers using the browser’s own encoder, with a main-thread fallback if workers are unavailable.
- Pages are generated from one registry: a tool is a directory, and routes, navigation, sitemap and internal links follow from it.
What it does not do
No online Photoshop, no layers, no AI background removal, no accounts, no template store, no video. The moment a photo editor is added, the 30 KB JavaScript budget and the “works offline” promise both die.
Known limitations
- AVIF can be read but not written — the browser’s canvas encoder cannot produce it.
- HEIC decoding depends on the browser; Safari does it, Chrome generally does not.
- Passport photos use explicit head-ratio guides rather than automatic face detection.