Arena, the AI leaderboard everyone uses, nearly doubles valuation to $3.1B
Lightspeed Venture Partners and Khosla Ventures led the $200M Series B, which lands about four months after Arena said it hit $100M in annualized run-rate revenue.
Arena, the crowdsourced leaderboard that began in 2023 as a research project at UC Berkeley, has raised a $200 million Series B at a $3.1 billion valuation, the company said Thursday.
The round was led by Lightspeed Venture Partners and Khosla Ventures, with Salesforce Ventures, 01 Advisors, Dell Technologies Capital, Endeavor Catalyst, a16z, Felicis and others joining in.
The valuation climb has been fast. Arena announced a $150 million Series A in January at a $1.7 billion post-money valuation, when its annualized revenue stood at $30 million. That means the company has nearly doubled its valuation in roughly 10 months. In June, Arena said it reached $100 million in annualized run-rate revenue.
The consumer side of the business is free: people enter prompts or request vibe-coded projects, then rate which model did it better. Arena claims tens of millions of monthly visitors.
Revenue comes from AI Evaluations, a commercial product introduced in September of last year that gives model labs and enterprises detailed performance analytics built on that community feedback. The timing proved fortuitous. This year, AI labs concluded that their models were gaming benchmarking tests, finding ways to rack up good scores without truly earning them. At the same time, enterprises wanted help figuring out which model worked best for their internal needs rather than relying only on standardized benchmarks.
"AI is advancing faster than our ability to evaluate it, and static benchmarks break down once models recognize they're being tested," the company said in its funding announcement. "The world needs a neutral third party to measure how safe and aligned AI actually is once it's in the hands of real people. Arena is stepping into that role today."
To that end, Arena added an alignment category to its leaderboard, ranking models on problems including unauthorized action, or taking actions a model wasn't asked to take; false attribution, or crediting statements or facts to the wrong source; and what Arena calls "deceptive completion," lying about finishing tasks it didn't do.
As it stands, a slate of OpenAI's models sit atop Arena's preliminary alignment leaderboard, with Claude Opus 5.5 and Claude Fable in sixth and ninth place, respectively.
The company did not disclose in the announcement how it intends to spend the new capital.