Chatbot Arena, hosted at lmarena.ai, is an open platform where users compare two anonymous AI chatbots side-by-side and vote for the better response, generating a live crowdsourced leaderboard of LLM quality. Developed by researchers at UC Berkeley, it provides an Elo-based ranking across dozens of models. Widely referenced by AI researchers and practitioners as a reliable model quality benchmark.
