Arena Raises $200M Series B at $3.1B Valuation to Rank AI Models and Agents

Arena, the AI evaluation platform that operated until recently as LMArena, has raised $200 million in a Series B round at a $3.1 billion valuation. Lightspeed Venture Partners and Khosla Ventures co-led the round.

Salesforce Ventures, 01 Advisors, Dell Technologies Capital and Endeavor Catalyst joined as new investors. Existing backers a16z, Felicis, AMP PBC, QuantumLight and The House Fund also took part.

What Arena Does

Arena runs one of the best-known public leaderboards for large language models. Instead of relying only on fixed benchmark tests, it asks real users to compare responses from competing models and vote on which is better. Those votes, collected at scale, produce rankings that model makers and enterprise buyers now watch closely.

The company has since expanded beyond chat. Agent Arena evaluates AI agents across full real-world workflows such as coding, document analysis and creative writing, rather than judging single answers. Arena argues that static benchmarks are losing value as models become capable enough to recognize when they are being tested.

The Numbers Behind the Valuation

Arena reports more than $100 million in annualized revenue. Its platform has logged roughly 350 million sessions and 62 million votes across text, vision, code, search, video and image, and it draws tens of millions of monthly visitors from more than 150 countries. Agent Arena alone recorded seven million sessions in under five months, according to Unite.AI.

The round follows a $100 million seed in May 2025 and a $150 million Series A in January 2026, meaning Arena has raised about $450 million in less than 18 months.

A New Alignment Index

Alongside the funding, Arena released a preview of its Alignment Index, which measures whether AI agents behave the way users intend. The index tracks three failure types: taking actions a user did not request, attributing statements or intentions to a user that the evidence contradicts, and reporting a task as finished when it is not. The preview draws on 90,000 real agent sessions across 27 models.

Arena plans to use the new capital to build out independent evaluation infrastructure for increasingly autonomous AI systems, with more safety signals, starting with how models handle harmful requests, to be added to the index.

Why It Matters

As AI labs compete for enterprise contracts, a neutral scorekeeper has become valuable in its own right. Arena's rankings already influence how companies market new models and how buyers choose between them. A $3.1 billion valuation shows investors believe evaluation itself can be a large business, and that trust in that score will matter more as AI agents take on real work with real consequences.

Source

The SaaS News — "Arena Raises $200M Series B"