LMArena Reaches $1.7B Valuation Following $150M Series A Round
LMArena, a startup that originally began as a research project at UC Berkeley in 2023, announced that it has raised $150 million in a Series A funding round, bringing its post-money valuation to $1.7 billion.
The round was led by Felicis and the university’s investment arm, UC Investments, with participation from leading investors including Andreessen Horowitz, Kleiner Perkins, and Lightspeed.
LMArena entered the commercial market with a $100 million seed round in May at a $600 million valuation, bringing its total funding to $250 million in under seven months, marking one of the fastest growth trajectories in the AI sector.
The company is best known for its crowdsourced AI model performance leaderboards, where users compare two models by submitting a prompt and selecting the better-performing result.
According to the company, the platform attracts more than 5 million monthly users across 150 countries, generating over 60 million interactions per month, making its rankings a key reference point for developers and enterprises.
LMArena evaluates a wide range of models, including OpenAI GPT, Google Gemini, Anthropic Claude, and Grok, as well as models specialized in image generation, text-to-image, coding, and computer vision.
Originally launched as Chatbot Arena, the open research project was founded by Anastasios Angelopoulos and Wei-Lin Chiang and funded initially through grants and donations.
As its leaderboards gained influence, LMArena partnered with leading AI model developers such as OpenAI, Google, and Anthropic to make their models available for interactive evaluation.
In September, the company launched its commercial offering, AI Evaluations, enabling enterprises, model labs, and developers to commission in-depth model assessments via the LMArena community. The service reached an annualized revenue run rate of $30 million within less than four months of launch.
