Text Behind Image

What they built

Text Behind Image does one thing well: it places text behind the subject of a photo, the effect you see on magazine covers and YouTube thumbnails. Doing it by hand in Photoshop is fiddly. Rexan Wong made it a single upload, and the site has now generated more than 600,000 designs.

Rexan is a self-taught high-school student from Hong Kong who was 16 when he launched it. He learned to code from a school elective and long-form YouTube tutorials, then built the whole thing with AI.

How they did it

He prototyped it in about three hours using Cursor and Vercel v0, with no formal design experience. The stack is Next.js, React, TypeScript, and Supabase on Vercel. Then he posted a demo video to X and went to bed. It hit hundreds of thousands of views inside 24 hours, went on to be Product of the Day on Product Hunt, and grew to 300,000 users within months, all with zero ad spend.

Monetization was reactive, not planned. An unexpected Vercel bill and an audience telling him to charge pushed him into a freemium model. He also open-sourced the whole project on GitHub, where it has thousands of stars, even while running the paid hosted version.

The numbers

• 300,000 users and about $1,000 MRR by February 2025, from ads, sponsorships, and 60 paid subscribers at $4 a month. source
• Roughly 5,000 site views a day at that point, with the product open-sourced under AGPL-3.0.
• The viral X traffic was a one-time surge; the recurring revenue is the monetization layered on afterward.

The revenue is modest and Rexan is honest about that. The real lesson is speed and distribution: a teenager shipped a genuinely useful tool in an afternoon with AI, and a single good demo did the marketing.