Enlarge an image 2× or 4× with an AI model that runs on your own machine.
The first run downloads the model your browser will execute — a Real-ESRGAN export. Your image is not part of that request, and the model is kept in your browser, so later runs — on later visits too — usually start from that copy instead of downloading it again. The model runs on the GPU where the browser offers one, and falls back to WebAssembly where it does not.
The source image can be up to 4 MP at x2 and 1 MP at x4 — the ceiling follows the size of the result, and both are lower than the one the other tools use. Still images only, and the result comes out as PNG.