We love investing at the moment of breakthrough — when bold research is ready to become a foundational company. There's a unique energy in that stretch between lab and launch, and our investment in LMArena is a perfect expression of what we built this firm to do.

LMArena started as a scrappy project out of UC Berkeley's LMSys research group, launching the now-famous Chatbot Arena leaderboard to bring science and transparency to LLM evaluation. In under a year, it's become the most widely used open benchmarking platform in AI, with over 3.5 million votes cast across 400+ model comparisons. But what they're building now is far more than a leaderboard. It's infrastructure.

The newly rebuilt platform at lmarena.ai is faster, cleaner, and designed mobile-first — with new features like Prompt-to-Leaderboard, which creates a custom leaderboard based on your query, and Style Control, which surfaces nuanced differences in how models behave. In a space where model development is outpacing evaluation, LMArena is stepping in with a neutral, reproducible, community-driven layer that helps us understand how models actually perform for real users, in real scenarios. This is the kind of work that shapes how our field moves forward and how it holds itself accountable.

The $100M seed round — led by a16z and UC Investments, with participation from Laude Ventures, Lightspeed, Felicis, Kleiner Perkins, and The House Fund — underscores how fundamental this platform is becoming. For us, this isn't just a product bet. It's a bet on the core infrastructure AI will rely on.

We've built companies with Ion Stoica before, and we're thrilled to partner with him again — this time alongside co-founders Anastasios Angelopoulos and Wei-Lin Chiang. Both deeply impressive technical founders with research roots, who we believe represent exactly the kind of leadership this field needs.

Laude exists to help technical founders turn discovery into enduring companies. LMArena is doing just that — placing science, transparency, and community participation at the heart of AI evaluation.

We're proud to be part of their journey. If you're a founder working on deep infrastructure to make AI more robust, fair, or usable — we'd love to hear from you: hello@laude.vc