LMArena lands $1.7B valuation four months after launching its product

LMArena, a startup that emerged from a research initiative at UC Berkeley in 2023, has announced a significant funding milestone. On Tuesday, the company revealed it had secured $150 million in a Series A funding round, bringing its post-money valuation to an impressive $1.7 billion. This funding round was spearheaded by Felicis and the university's investment arm, UC Investments. The company made a strong entrance into the commercial sector earlier this year with a $100 million seed round in May, which established its initial valuation at $600 million. With the latest funding, LMArena has successfully raised a total of $250 million within a mere seven months. Renowned for its crowdsourced AI model performance leaderboards, LMArena offers a unique platform where users can test various AI models by submitting prompts and selecting which model performs better. Currently, the platform boasts over 5 million monthly users from 150 countries, generating around 60 million conversations each month. This vast user engagement directly contributes to the creation of their comprehensive leaderboards, which evaluate models based on tasks ranging from text generation to web development and beyond. LMArena's testing includes a diverse array of models, such as OpenAI's GPT, Google Gemini, Anthropic's Claude, and Grok, alongside specialized models designed for tasks like image generation and advanced reasoning. Initially known as Chatbot Arena, the project was started by UC Berkeley researchers Anastasios Angelopoulos and Wei-Lin Chiang, funded through grants and donations. As LMArena shifted gears towards revenue generation, it established partnerships with notable AI companies, including OpenAI, Google, and Anthropic, allowing these firms' flagship models to be evaluated by LMArena's community. However, this shift hasn't been without controversy; in April, a group of competitors accused LMArena of enabling model makers to manipulate benchmark results, claims that the startup has robustly denied. With the launch of its commercial service, AI Evaluations, in September, LMArena began offering enterprises and developers the opportunity to evaluate AI models through its community, which has led to an annual recurring revenue (ARR) of $30 million within just four months of its official launch. This rapid growth and the company's increasing popularity have attracted significant interest from venture capitalists, culminating in the recent Series A funding that also saw participation from renowned firms like Andreessen Horowitz and Kleiner Perkins.