The page is dated October 22, 2026 and reports 923,854 votes.
Kimi K3.1 Pro has a preliminary score of 1692, with an uncertainty range of +14/-14 and a displayed rank spread of 7 to 13. Anthropic’s Claude Opus 5.5 (Max) leads at 1807, followed by OpenAI’s GPT-6 Astra (Max) at 1774 and Meta’s Muse Wonder Full (Max) at 1726. Google’s Gemini 4 Argon (High) is fifth at 1674.
Arena describes the ranking as covering front-end web-development tasks, including agentic coding workflows that require multi-step reasoning and tool use. The listing identifies Kimi K3.1 Pro as proprietary and marks its result preliminary.
The leaderboard does not establish Kimi K3.1 Pro’s price, availability, release terms, or performance on coding work outside Arena’s evaluation setup. It also does not establish a causal comparison with the models ranked above or below it.