a queue based image resizing tool, built for speed ⚡️

- bun – runtime (blazingly fast)
- hono – backend framework
- bullmq – job queue
- sharp – image resizing + webp conversion
- redis – job storage
- react – frontend
you upload an image → it gets queued → resized into multiple sizes → saved as .webp
→ served to you via urls.
- upload – image hits the backend
- queue – job added to bullmq
- process – sharp resizes it in the background
- store – saved to
public/images/<jobId>
- respond – returns urls of resized images




you probably dont even need to read any further - becuz images tell you everything
- multiple sizes – customize resize steps (90px, 120px, etc.)
- async processing – doesn't block anything
- served as static – use the URLs anywhere
- bun makes it ridiculously fast
- sharp is memory efficient
requires macOS (uses brew
in entrypoint.sh
, tweak for other OS)
before running, make sure you install deps:
cd ui && bun install
cd ../backend && bun install
then start it up:
./entrypoint.sh
spins up redis → backend → frontend
you’re live. go throw some images in! 😎