According to the July 10, 2026 update to the LM Arena text leaderboard, Anthropic's Claude-Fable-5 is maintaining its lead position with a score of 1505 and a confidence interval of ±8, backed by 7,252 votes. The text arena includes rankings for 374 models, with data drawn from over 7.27 million total votes accumulated across the entire platform. The leaderboard operates through repeated pairwise comparisons where users evaluate responses from two models without knowing their identities, with each vote contributing to an overall score reflecting perceived quality across criteria such as coherence, relevance, and depth of answers. The July 10 snapshot records the state after sufficient comparisons have occurred for the listed entries, serving as a reference point rather than a final or unchanging order.