CathieDaDa/Reasoning_Capability_Leaderboard

1排名,模型名称,综合加权得分,常识推理,学科推理,不确定性下的决策推理,道德与伦理推理21,Gemini 2.5 Flash,92,98,93,89,8732,豆包1.5 Pro（思考模式）,91,97,92,88,8742,Gemini 2.5 Pro,91,93,94,90,8754,Grok 3（思考模式）,90,96,88,89,8665,GPT-5（自动模式）,89,88,98,88,8375,混元-T1,89,97,95,84,8185,通义千问3（思考模式）,89,96,89,86,8595,文心一言 X1-Turbo,89,98,85,86,86109,DeepSeek-R1,87,94,93,78,82119,通义千问3,87,97,79,87,86129,文心一言4.5-Turbo,87,96,76,87,871312,混元-TurboS,86,96,79,83,841413,豆包1.5 Pro,85,97,81,86,741513,GPT-4.1,85,97,70,87,861613,GPT-o3,85,90,95,73,801713,Grok 3,85,97,69,87,861813,Grok 4,85,82,87,82,871917,DeepSeek-V3,84,95,81,84,772019,GPT-4o,82,98,65,87,782119,GPT-o4 mini,82,91,87,72,762221,Claude 4 Opus（思考模式）,81,96,84,72,712321,MiniMax-01,81,96,69,83,752421,360智脑2-o1,81,93,76,81,722524,Claude 4 Opus,80,95,85,70,702624,GLM-4-plus,80,93,71,83,732724,Step 2,80,97,63,82,782827,Yi- Lightning,79,97,59,82,792927,Kimi,79,94,61,79,813029,Spark 4.0 Ultra,78,91,71,75,763130,日日新 V6 Pro,77,86,58,84,783231,GLM-Z1-Air,76,90,76,73,643332,Llama 3.3 70B,75,82,52,83,813433,日日新 V6推理,74,96,63,68,703534,Baichuan4-Turbo,71,91,48,77,693635,Step R1-V-Mini,66,96,80,37,513736,Kimi-k1.5,66,84,79,42,58