中国开源模型Kimi K2.6在编程挑战中击败Claude、GPT-5.5和Gemini
在最近的AI编程挑战赛中，中国初创公司Moonshot AI开发的开源模型Kimi K2.6 以22分胜出，领先于小米的MiMo V2-Pro，并超越了OpenAI的GPT-5.5和Anthropic的Claude Opus 4.7。
该挑战赛为“单词宝石拼图”，模拟滑动字母拼图游戏，测试模型的实时编程能力。
Kimi K2.6 凭借其“贪婪式滑动”策略，即使在较大网格中也取得了领先，展现了开源模型在特定任务上的竞争力。
此次挑战突显了开源模型追赶前沿模型的趋势，也表明模型性能差异在缩小。
查看原文开头（英文 · 仅前 3 段）
By Rohana Rezel
I’m running the ongoing AI Coding Contest where I pit major language models against each other in real-time programming tasks with objective scoring. Day 12 was the Word Gem Puzzle. Ten models entered. The results were not what most people would have predicted.
Kimi K2.6, an open-weights model from Chinese startup Moonshot AI, won the challenge outright: 22 match points, 7-1-0. MiMo V2-Pro from Xiaomi came second. GPT-5.5 was third. Claude Opus 4.7 finished fifth. Every model from the Western frontier labs landed below the top two.
※ 出于版权考虑，仅引用前 3 段。完整内容请阅读原文。