速递｜从学术项目到17亿美元估值：LMArena凭“模型竞技场”七个月内估值翻近三倍

Core Insights - LMArena, a startup initially established as a research project at the University of California, Berkeley, announced a successful completion of a $150 million Series A funding round, achieving a post-money valuation of $1.7 billion [1] - The company has rapidly evolved from a research initiative to a commercial entity, raising a total of $250 million in approximately seven months, following a $100 million seed round in May with a valuation of $600 million [2] Group 1 - LMArena is known for its crowdsourced AI model performance leaderboard, allowing users to input prompts and select the better-performing model from two options, with over 5 million monthly users across 150 countries generating 60 million dialogues each month [2] - The platform ranks various models based on multiple task dimensions, including text processing, web development, visual recognition, and text-to-image generation, featuring models from OpenAI, Google, Anthropic, and others [2][3] - The company transitioned from its original identity as Chatbot Arena, founded by researchers Anastasios Angelopoulos and Wei-Lin Chiang, and initially funded through research grants and donations [3] Group 2 - LMArena's leaderboard has garnered significant attention from model developers, and the company has partnered with leading model firms like OpenAI and Google for community evaluations [3] - In September, LMArena launched a commercial AI evaluation platform, generating an annual recurring revenue (ARR) of $30 million within just four months of service launch [3] - The growth trajectory and popularity of the startup attracted numerous venture capital firms to participate in its Series A funding round, including Andreessen Horowitz, Kleiner Perkins, and Lightspeed Venture Partners [4]