The AI model leaderboard platform Arena, which began as a research project at UC Berkeley in 2023, has hit $100 million in annualized run-rate revenue just eight months after launching its commercial service.
Arena is best known for its crowdsourced AI model performance leaderboard, built from over 10 million user evaluations. Users type a prompt and compare outputs from two anonymous models, voting on which performs better. The public leaderboard remains free, but Arena started generating revenue in September with the launch of AI Evaluations, a service offering deep-dive performance analytics to model labs and enterprises.
“A lot of people don’t even understand that our business is making any money at all; people still see us as like an open-source project,” co-founder and CEO Anastasios Angelopoulos told TechCrunch.
The company competes indirectly with human-labeling startups like Mercor, Surge, and Scale AI, which help model makers refine AI during post-training. As AI providers race to maximize performance, demand for these refinement services continues growing rapidly. When Arena raised a $150 million Series A in January at a $1.7 billion valuation, its annualized revenue was $30 million. The jump to $100 million underscores the exploding appetite for model evaluation.
Founded by Angelopoulos, CTO Wei-Lin Chiang, and UC Berkeley professor Ion Stoica (a Databricks co-founder), Arena has raised $250 million total from investors including Felicis, Andreessen Horowitz, Kleiner Perkins, Lightspeed Venture Partners, and others. The platform ranks models across text, coding, vision, image generation, and complex workflows through its recently introduced Agent Mode.
Source: TechCrunch
