You describe a rainbow umbrella to an AI. It scores how well your prompt matches a reference image. Top performers hit 88% accuracy on this exact challenge.
Prompt Challenge turns prompt engineering into a competitive game with 71 challenges across four difficulty levels. You'll face two types of tasks. Describing images so AI can recreate them. Prompting AI to build specific websites. Each attempt gets scored against reference materials.
Daily challenges keep things fresh. A leaderboard tracks your progress against 29 other players currently working through prompts. Easy challenges might ask you to describe simple objects — Expert level tests complex scenarios.
Marketing managers launching AI initiatives would find this useful for building practical prompting skills before deploying tools company-wide.
Scoring measures precision. If you prompt "create a website for a bakery" and the AI output closely matches their reference bakery site, you score high. Miss key details and your score drops. This isn't about creativity. It's about control and consistency.
Four difficulty tiers mean you won't get overwhelmed starting out. It doesn't offer much beyond the core challenge format though. No tutorials or explanations help you understand why certain prompts work better than others.
