DesignArena raises $7.9M to help AI models learn what humans actually want

DesignArena, a platform that gathers human preferences on AI-generated creative outputs, has secured $7.9 million in seed funding to expand its role as a critical evaluation tool for the AI industry. Index Ventures led the round, with backing from Conviction, A*, Valkyrie, and others.

The company addresses a real problem: while AI models can now produce functional designs, websites, and images, determining whether those outputs are actually good requires human judgment at scale. DesignArena built a two-sided marketplace to solve this gap—offering end users a straightforward interface to compare and rank AI outputs, while selling the resulting preference data to frontier AI labs building generative models.

How it works for users#

The platform operates much like a sophisticated routing tool. Users enter a design prompt, select their desired format (websites, images, or other visual media), and specify style preferences. The system then presents a series of side-by-side comparisons—essentially A vs. B decisions—until the user has ranked several AI-generated options from best to worst.

For individual users, this is a free way to access and refine outputs from multiple AI models in one place. But the real business emerges elsewhere.

The enterprise opportunity#

Frontier AI labs—companies building the underlying generative models—have become eager customers. By participating in DesignArena, they gain access to continuous, real-world feedback on what users prefer, allowing them to improve their creative capabilities in ways that automated benchmarks cannot easily capture.

"It was the missing bottleneck for a lot of these models to make improvements in the design space," co-founder Grace Li explained. The company reportedly closed its first major deal with a frontier lab within a week of identifying the opportunity and has since scaled to 5.3 million users globally.

That demand is reflected in the financials: Intelligence (the company behind DesignArena) is currently generating $60 million in annual recurring revenue, cementing its place as an essential source of human-led evaluation data for the industry.

The data advantage#

A key competitive asset is the platform's ability to track user preferences across geographies and time. Because users must log in to retrieve their outputs, Intelligence can observe how design tastes vary—for instance, noting that web dashboards in Asia tend toward more maximalist styles. These insights complement automated benchmarks, which, while scalable, can be gamed or manipulated, as underscored.