Design Arena: $7.9M to Boost AI Valuation

⚡

Key Takeaways

1Design Arena, founded by Grace Li and her colleagues, has raised $7.9 million to enhance AI models.

2The tool is used by 5.3 million people and generates $60 million in annual recurring revenue.

3Users rank the model results, providing critical feedback to AI companies.

💡Why it matters — Design Arena fills a crucial gap by providing large-scale human feedback, which is essential for improving AI models.

A Promising University Genesis

Grace Li, co-founder of Design Arena, launched her company with a group of friends just before their graduation in 2025. Their initial ambition was to develop a game engine based on artificial intelligence. Although the models they created could produce functional games, they severely lacked entertainment value. This raised a fundamental question: how do you determine if a game is fun?

To answer this question, they concluded that no algorithm could replace human judgment. They began to brainstorm ways to obtain honest and large-scale feedback from users. This is how Design Arena was born, an AI tool that has captivated 5.3 million users worldwide. Many AI companies saw this tool as a solution for obtaining user feedback at scale, and many were willing to pay for it.

Rapid Commercial Success

Li explains that Design Arena filled a crucial gap for many AI models in terms of design. In just one week, the company signed its first major contract with a leading lab, marking the beginning of a success story. Recently, the parent company of Design Arena, named Intelligence, announced that it had raised $7.9 million in a seed funding round. This round was led by Index Ventures, with participation from Conviction (Sarah Guo and Mike Vernal), A*, Valkyrie, and other investors.

An Innovative User Interface

For individual users, Design Arena functions like a sophisticated model router. The interface is similar to that of ChatGPT, featuring a prompt window and dropdown menus for different visual formats such as websites and images. After submitting a request, the user is prompted to choose between several "A versus B" options to rank the results from best to worst.

However, the true value of the platform lies in its use by companies. These businesses can leverage Design Arena to obtain instant and endless feedback on their media-generating models. Users, often indifferent to the models they are evaluating, simply seek the best possible result, thus providing valuable insights into user preferences.

A Significant Revenue Source

For leading labs, this service is valuable enough to justify an investment. Design Arena currently generates $60 million in annual recurring revenue (ARR), solidifying its position as an essential source of human evaluation data for the AI industry.

Users must log in to access their results, allowing Intelligence to track the evolution of tastes across different continents and over time. For example, web dashboards in Asia tend to adopt a more maximalist design style. This data complements automated benchmarks, which, while effective at scale, can be manipulated, as demonstrated by the recent data breach at Hugging Face.

An Evolving Market

However, the success of crowdsourced human feedback is not guaranteed. For instance, Yupp, a similar company, shut down less than a year after its launch. It had raised $33 million from Chris Dixon of a16z crypto and boasted over 1.3 million users. In contrast, other startups like LM Arena are thriving. The latter, which focuses on text responses, raised $150 million in a Series A funding round just four months after launching its paid product.