Arena, the AI model leaderboard that originated as a research project at UC Berkeley in 2023, has reached $100 million in annualized run-rate revenue — just eight months after launching its commercial service in September 2025.
The startup is best known for its free, crowdsourced leaderboard that ranks AI models based on more than 10 million user evaluations. Users submit a prompt that is sent to two competing models simultaneously, then vote on which response is better. While that leaderboard remains free, Arena began monetizing in September 2025 with the launch of AI Evaluations, a paid service that gives model labs and enterprises detailed performance analytics drawn from its user community.
Arena’s annualized revenue stood at $30 million when the company announced a $150 million Series A in January 2026, at a post-money valuation of $1.7 billion. The jump to $100 million came within months. CEO Anastasios Angelopoulos noted that the company’s commercial activity is still widely underestimated. “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,” he told TechCrunch. He also clarified that despite using the term ARR, Arena’s revenue is consumption-based rather than strictly recurring.
Arena was co-founded by Angelopoulos and Wei-Lin Chiang, both UC Berkeley postdoctoral students, along with Ion Stoica — a UC Berkeley professor and Databricks co-founder — who advised the project before it incorporated as a company in April 2025. The startup has raised a total of $250 million from investors including Andreessen Horowitz, Kleiner Perkins, Lightspeed Venture Partners, and Felicis, among others.
The company competes for enterprise budgets against human labeling firms such as Scale AI, Mercor, and Surge, all of which serve AI model makers during post-training. Arena’s platform covers text, coding, vision, and image generation tasks, and recently added Agent Mode for complex, long-running workflows. The rapid revenue growth suggests that demand for third-party model evaluation services continues to expand as AI providers invest heavily in post-training refinement.
Source: TechCrunch