Arena, the startup behind a crowdsourced AI model leaderboard, says it has raised $200 million in a Series B round at a $3.1 billion valuation. The company reported a $100 million annualized revenue run rate in June.
The platform lets people compare AI systems by entering prompts and voting on which model performs better. Arena says it attracts tens of millions of visitors each month.
Last September, the company introduced AI Evaluations, a paid product for model developers and enterprise customers. It offers detailed performance analytics based on feedback from Arena’s community. TechCrunch reports that Arena began in 2023 as a UC Berkeley research project crowdsourcing ratings of AI models.
Arena raised $150 million in a Series A round in January at a post-money valuation of $1.7 billion. Its annualized revenue run rate was then $30 million, according to the company. The new valuation is nearly double the January figure.
Lightspeed Venture Partners and Khosla Ventures co-led the Series B. Salesforce Ventures, 01 Advisors, Dell Technologies Capital, Endeavor Catalyst, Andreessen Horowitz, Felicis and other investors also participated.
Arena also added an alignment category to its leaderboard. It evaluates whether models take actions without being asked, attribute information to the wrong source, or claim to have completed tasks they did not actually perform.
Several OpenAI models currently top the preliminary alignment rankings. Claude Opus 5.5 and Claude Fable are in sixth and ninth place, respectively.