There’s this AI image generation platform that one guy built solo over a whole year, and it just started closed beta. Honestly, it hit me pretty hard. OP started coding it in his spare time back when AI image gen first blew up in late 2023, originally just for himself and his friends—wanted to move his workflow across different devices. But it kept growing, and now he’s opened it up for beta testing. The site’s called vikify.io.
It’s built on ComfyUI’s API underneath, aiming to give people who want to create freely fewer restrictions. That said, prompts still get classified by AI and blocked when necessary for compliance—claims no logs are kept. Features-wise, it’s loaded with over a hundred LoRAs, more than twenty samplers, and advanced controls like seed, CFG, aspect ratio are all open for tweaking.
For a product solo-developed over a year, the polish is pretty impressive. What I’m curious about is how they’ll sustain the costs going forward, since image gen burns through compute power like crazy. Any bros in the beta care to share your experience?