LMArena
A public chat-preference evaluation surface with source-defined preference ratings and model comparisons.
What this benchmark measures
A public chat-preference evaluation surface with source-defined preference ratings and model comparisons.
Rows on this page are sourced from public benchmark artifacts, leaderboard exports, or source-linked model reports. Each row keeps benchmark version, source model name, and available run details attached to the score.
The metric shown here is source-defined rating. It should be interpreted within LMArena, not compared as part of a site-wide ranking.
What to be careful about
Ranking-native UX conflicts with evals.report tone. Include only with careful framing.
Frequently asked
What is LMArena?
A public chat-preference evaluation surface with source-defined preference ratings and model comparisons. It is a chat preference benchmark measured by source-defined rating.
What does source-defined rating mean on LMArena?
LMArena reports source-defined rating; higher is better. Scores are shown only within LMArena and are never averaged with other benchmarks.
What is the top reported LMArena score?
Claude Opus 4.6 has the top reported score on LMArena: 1499 (source-defined rating).
Why do LMArena scores differ across runs?
Harness, scaffold, reasoning effort, and prompt setup change results, so two runs of the same model can differ. evals.report keeps each score with its run context so the differences stay visible.
Does evals.report rank models across benchmarks?
No. LMArena scores are shown within their own metric; evals.report never combines benchmarks into a composite ranking or a single "best model".