Fix AI-generated pixel art and vector images right in your browser.
Bridge the gap between raw AI generation and clean, usable graphic assets.
unfake.js is a browser-based library and tool designed to process AI-generated pixel art and raster images into clean, usable assets. It is ideal for graphic designers, pixel artists, and creative technologists working heavily with text-to-image models.
The library offers a Pixel Art Processor with intelligent scale detection, content-aware downscaling, and grid snapping, alongside an Image Vectorizer powered by imagetracer.js and WASM pre-filters. It runs in the browser via an interactive interface built with Tweakpane or as an ES module in your own JavaScript pipeline.
It suits graphic designers and developers struggling with AI pixel art artifacts, uneven grids, and messy traces. It does not suit users looking for a full-featured standalone vector illustration suite like Illustrator.
Automatically detects true pixel sizes or fractional grids from upscaled AI images.
Reduces images to native 1x scale without introducing new colors or blurriness.
Limits color palettes using imagequant in a WASM core or custom fixed palettes.
Converts cleaned rasters into scalable SVGs using pre-processing filters and imagetracer.js.
Free under its open-source license. There is no hosted paid option mentioned, so the only cost is your own hardware and time to run it.
880 open remote AI-training roles, updated November 15, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
These reviewed AI tools do the same job instantly — from our scored directory.
Yes, you can import processImage and vectorizeImage directly as ES modules in your own web applications.
Modern browsers restrict ES module imports from local file:// paths, requiring a local http(s) server to load the WASM and scripts correctly.
No, all image processing runs entirely client-side inside your browser via WebAssembly.
See Zekai first in Google