跳过主要内容
评测来源/LiveBench 推理与数学
观察中推理

LiveBench 推理与数学

LiveBench · LiveBench 推理与数学 的公开模型评测结果。

官方评测
来源说明

该来源测量模型在「推理」领域的具体交付表现,原始成绩会和统一能力读数分开展示。

限制: 不同版本、推理设置和测试集可能导致分数变化,请结合数据时间阅读。

使用方式

参与排名的来源会影响能力分;参考、交叉参考和观察来源仅用于对比与核验。

本来源的模型成绩

GPT-6.1 Sol94.73312594.72026-06-25
GPT-6 Astra94.7394.72026-06-25
Claude Opus 5.594.61562594.62026-06-25
Claude Fable 5.194.349594.32026-06-25
GPT-5.6 Sol93.9257593.92026-06-25
Claude Sonnet 5.593.88312593.92026-06-25
Muse Wonder93.552593.62026-06-25
Claude Opus 593.4712593.52026-06-25
Claude Fable 592.8197592.82026-06-25
Muse Spark 1.392.80187592.82026-06-25
GPT-5.6 Terra92.77137592.82026-06-25
GPT-5.592.755592.82026-06-25
GPT-6 Sol92.50487592.52026-06-25
Claude Opus 4.891.75437591.82026-06-25
Grok 4.691.5387591.52026-06-25
GPT-5.491.1317591.12026-06-25
Claude Sonnet 590.8172590.82026-06-25
Muse Spark 1.290.60412590.62026-06-25
DeepSeek V4 Pro 081390.4637590.52026-06-25
Gemini 3.8 Flash90.4287590.42026-06-25
Claude Opus 4.790.02390.02026-06-25
DeepSeek V4.1 Flash89.99190.02026-06-25
Qwen3.8 Max89.76187589.82026-06-25
Grok 4.789.16587589.22026-06-25
Grok 4.588.9987589.02026-06-25
Claude Opus 4.688.99589.02026-06-25
Mistral Large 488.763588.82026-06-25
GPT-5.288.1887588.22026-06-25
Kimi K387.55487587.62026-06-25
Gemini 3.1 Pro Preview87.52487587.52026-06-25
Muse Spark 1.187.4377587.42026-06-25
Claude Haiku 5.586.99162587.02026-06-25
GLM 5.386.850586.92026-06-25
DeepSeek V4 Flash 073186.71212586.72026-06-25
DeepSeek V4 Pro86.68387586.72026-06-25
GPT-5.6 Luna86.422586.42026-06-25
Claude Sonnet 4.685.88162585.92026-06-25
Gemini 3.6 Flash85.7757585.82026-06-25
GPT-6 Luna85.44637585.42026-06-25
Claude Opus 4.585.2377585.22026-06-25
Qwen3.7 Max84.29237584.32026-06-25
GLM-5.284.20312584.22026-06-25
Inkling83.355583.42026-06-25
Kimi K2.681.82562581.82026-06-25
Nemotron 3 Ultra 550B A55B81.68012581.72026-06-25
Kimi K2.7 Code81.20681.22026-06-25
Qwen3.6 Plus79.77587579.82026-06-25
GLM-5.3 Flash79.44437579.42026-06-25
Grok 4.377.580577.62026-06-25
Grok Build 0.177.399577.42026-06-25
MiniMax-M375.71412575.72026-06-25
DeepSeek V4 Flash75.11462575.12026-06-25
Qwen3.6 27B75.07612575.12026-06-25
GPT-5.4 mini74.89287574.92026-06-25
Gemini 3.5 Flash Lite66.963567.02026-06-25

常见问题