Arena, the AI model leaderboard that originated as a UC Berkeley research project in 2023, disclosed on June 29, 2026 that it has reached $100 million in annualized run-rate revenue — tripling from $30 million in January — just eight months after launching AI Evaluations, its first paid product. Arena's free consumer leaderboard, generated from more than 10 million crowdsourced user comparisons between AI models, remains the most widely cited independent ranking in the industry; the monetization comes from selling model labs and enterprises deep-dive performance analytics built on that same evaluation data.
The growth trajectory is unusually steep for an AI-infrastructure-adjacent business: Arena raised a $150 million Series A in January at a $1.7 billion post-money valuation when its ARR was $30 million, and has since more than tripled revenue without needing to raise again. Total capital raised stands at $250 million from investors including Felicis, Andreessen Horowitz, The House Fund, Kleiner Perkins, Lightspeed and UC Investments.
“Total capital raised stands at $250 million from investors including Felicis, Andreessen Horowitz, The House Fund, Kleiner Perkins, Lightspeed and UC Investments.”
The business model raises an inherent tension that Arena will have to keep navigating publicly: it sells evaluation services to the same model labs (OpenAI, Anthropic, Google, and others) whose products it ranks for free on its public leaderboard. Critics have flagged the potential conflict of interest — paying customers gaining insight into evaluation methodology that free users don't see — as the central risk to Arena's credibility as an independent benchmark.
Competitive context: Scale AI built a large evaluation and data-labeling business before its Meta-driven restructuring; Hugging Face runs its own open leaderboards without a comparable paid tier; smaller players like Vals AI and Patronus AI compete in enterprise-focused eval niches. Arena's advantage is scale and brand trust built over three years of free community use, which is difficult for a paid-first competitor to replicate.
What to watch: whether Arena maintains a credible separation between its free public leaderboard and paid evaluation business as scrutiny increases, whether it raises a growth round at a valuation reflecting the new $100M ARR, and how model labs respond if Arena's rankings start to meaningfully move enterprise purchasing decisions.