Arena Raises $200M Series B at $3.1B Valuation
Arena, the crowdsourced AI model leaderboard that began in 2023 as a research project at UC Berkeley, has raised a $200 million Series B round at a $3.1 billion valuation, the company announced Thursday. The new valuation comes just ten months after Arena closed a $150 million Series A at a $1.7 billion post-money valuation in January 2026.
The round was led by Lightspeed Venture Partners and Khosla Ventures, with participation from Salesforce Ventures, 01 Advisors, Dell Technologies Capital, Endeavor Catalyst, a16z, Felicis, and others.
From $30M to $100M in Annualized Revenue
Arena's rapid valuation jump tracks its equally rapid revenue growth. At the time of its Series A in January 2026, Arena reported $30 million in annualized run-rate revenue. By June 2026, that figure had reached $100 million.
The company offers a free, crowdsourced platform where users submit prompts or request vibe-coded projects and then rate which AI model performs better. Arena claims tens of millions of monthly visitors.
AI Evaluations: Commercializing the Leaderboard
In September 2025, Arena launched its commercial product, AI Evaluations, which provides model labs and enterprises with detailed performance analytics derived from community feedback.
The timing proved impeccable. In 2026, AI labs began realizing that their models were gaming benchmarking tests—finding ways to accumulate high scores without truly earning them. At the same time, enterprises sought help determining which models best suited their internal needs rather than relying solely 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," Arena 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."
New Alignment Leaderboard
To that end, Arena has added a new category to its leaderboard: alignment. This ranking evaluates models based on issues including:
- Unauthorized action: taking actions the model wasn't asked to take
- False attribution: wrongly crediting statements or facts to the wrong source
- Deceptive completion: lying about completing tasks that weren't done
Currently, a slate of OpenAI models tops Arena's preliminary alignment leaderboard, with Claude Opus 5.5 and Claude Fable in sixth and ninth place, respectively.
Why It Matters
Arena's rise reflects a broader shift in the AI industry: as models grow more capable, traditional static benchmarks are proving insufficient. A neutral, community-driven evaluation platform—one that measures real-world performance and alignment rather than test-taking ability—has become increasingly valuable to both AI labs and enterprises navigating a crowded model landscape.