We built imageTprompt because writing a great AI art prompt is hard — and you shouldn't have to spend 20 minutes crafting one when you already have a visual idea in your head.
Bridge the gap between visual inspiration and AI art generation. Whether you're a digital artist, a designer, or someone who just discovered Midjourney — you deserve a tool that translates your vision into words an AI can understand, instantly.
It started with a simple frustration: staring at an image on Pinterest, knowing exactly what aesthetic we wanted, but not knowing the right words to feed a text-to-image model. We tried describing it manually — and the results were always wrong.
So we built imageTprompt: a tool that uses multimodal AI to look at an image and generate the prompt that would recreate it. We open-sourced the analysis approach and built a clean, fast web interface around it.
Today, imageTprompt is used by artists, designers, content creators, and AI enthusiasts who want to spend less time on prompt engineering and more time creating.
Next.js 16
App Router, Turbopack, React 19
Cloudflare Workers AI
LLaVA 1.5 7B vision-language model
Tailwind CSS v4
Utility-first styling
TypeScript
Strict mode throughout
Founder & Developer
Building everything, slowly.
You?
We're always open to contributors.
Want to contribute or collaborate? Get in touch