Product
Rendragon
Create stunning AI art in seconds.
An AI image and video studio: 11 image models and 7 video models, upscaling, background removal and a canvas for editing, with a library, a community and credits.
Visit live site (opens in a new tab)- Type
- Product
- Role
- Design & development, front end and back end
- Disciplines
- Product Design, Full-stack, AI, UX/UI
Overview
Rendragon makes images and video with the current generation of models (Flux 2, Ideogram v3, Seedream 4, Nano Banana, Veo 3.1, Kling 2.5, Wan 2.5 and more), upscales them, removes backgrounds, guides a picture by another’s edges, depth or pose, edits them on a canvas by mask, and keeps them in nested collections, with a community feed, likes and follows.
I designed and built it, front end and back end, in Next.js 16 and TypeScript: PostgreSQL with Drizzle, Clerk for accounts, Cloudflare R2 for the media, Stripe for credits and five subscription tiers, and rate limits and daily spending caps by tier. Each generation runs on Replicate and comes back by a signed webhook; credits are taken only once it succeeds, and a scheduled job recovers any whose webhook never arrives.