Evaluation in real use
The platform collects human feedback on outputs from different AI models, giving developers and users a way to compare performance across tasks. Its community generated more than 60 million conversations a month at the announcement.
A commercial service
LMArena launched paid AI Evaluations in September 2025. Customers include model labs and enterprises seeking feedback on software engineering, scientific research and other practical use cases.
Where the funding goes
- Operate the evaluation platform, expand the technical team and strengthen research capabilities.