Arena AI Leaderboard Reaches $100M Annual Revenue
Original: Arena, the AI leaderboard everyone uses, is now a $100M business
Why This Matters
Arena's rapid revenue growth demonstrates strong market demand for AI model evaluation services as developers prioritize post-training optimization and model performance refinement.
Arena, a UC Berkeley-originated AI model leaderboard platform, achieved $100 million in annualized run-rate revenue just eight months after launching commercial services in September 2025, based on crowdsourced evaluations from over 10 million users.
Arena, which began as a research project at UC Berkeley in 2023, has grown into a significant commercial venture. The platform's free leaderboard service crowdsources AI model performance evaluations, with users comparing model outputs and selecting better responses. The company monetized through AI Evaluations, a service launched in September 2025 that provides detailed performance analytics to model laboratories and enterprises. The revenue represents consumption-based pricing rather than traditional recurring revenue. Arena's annualized revenue jumped from $30 million in January 2026 to $100 million by June 2026. Co-founder and CEO Anastasios Angelopoulos noted that many people still view Arena as an open-source project rather than a revenue-generating business. The company competes in the post-training refinement market against firms like Scale AI, Surge, and Mercor, which help model creators improve AI during development. Arena ranks models across text, coding, vision, image generation, and complex workflows through its Agent Mode. The startup was co-founded by Angelopoulos, UC Berkeley postdoctoral researcher Wei-Lin Chiang (CTO), and professor Ion Stoica, a Databricks co-founder. In January 2026, Arena raised a $150 million Series A at a $1.7 billion post-money valuation.