LMArena Hits US$1.7 Billion Valuation After Series A Round

AI evaluation startup LMArena reached a US$1.7 billion valuation after securing US$150 million in a Series A funding round. This milestone represents a three-fold increase in market value over an eight-month period, establishing the platform as a primary benchmark for large language models.

"To measure the real utility of AI, we need to put it in the hands of real users. LMArena does exactly this," says Anastasios Angelopoulos, Co-Founder and CEO, LMArena.

The Generative AI sector has experienced an unprecedented flow of capital since the launch of ChatGPT in 2022. This surge has driven demand for objective metrics to compare the performance of various systems. LMArena, originally established as Chatbot Arena by researchers at the University of California, Berkeley (UC Berkeley), has transitioned from an academic initiative to a commercial entity critical for technological evaluation infrastructure.

The operational model of the company differs through its implementation of blind, crowdsourced evaluations. Unlike traditional static benchmarks, which are often vulnerable to data contamination from model training, LMArena employs a direct comparative system where users rate responses from anonymous models in real time. This approach provides a dynamic performance metric that reflects practical utility in tasks such as text generation, web development, and complex reasoning.

Felicis and UC Investments led the recent funding round, with participation from venture capital companies including Andreessen Horowitz, Kleiner Perkins, and Lightspeed Venture Partners. Including the US$100 million seed round obtained in May, LMArena has raised a total of US$250 million within seven months.

Since its formal commercial launch four months ago, the company has reported significant growth in:

• Traffic Volume: The platform exceeds five million monthly users from 150 countries.

• Data Interaction: The system generates about 60 million monthly conversations to power its leaderboards.

• Financial Performance: LMArena reached an annualized revenue run rate of US$30 million as of December 2025.

LMArena expanded its technical offering through "AI Evaluations," a commercial service designed for AI labs and developers to request specific model assessments before mass deployment. This monetization strategy relies on strategic partnerships with leading developers such as OpenAI, Google, and Anthropic. Those companies regularly submit models — such as GPT-4o, Gemini, and Claude — to compete within the ecosystem of the platform.

According to UC Berkeley’s research, the competitive advantage of LMArena lies in the application of the Elo rating system to rank the relative capability of large language models. This methodological framework allows for rigorous technical comparison between proprietary and open-source models, while eliminating biases inherent in automated laboratory testing.

The company will use the capital to expand its technical team, strengthen research capabilities, and operate its scalable infrastructure. As the industry moves toward multimodal models that integrate computer vision and text-to-image conversion, LMArena intends to standardize the evaluation of these technologies.