LiveBench 语言模型排行榜

按 LiveBench 综合分比较模型;先计算各类别平均分,再计算综合分。

评测来源 LiveBench采集 / 更新 查看原榜评分方法

LiveBench 综合分

64 个模型
LiveBench · LiveBench 综合分
排名模型 / 开发机构LiveBench 综合分推理编程智能体编程详情
01Claude Fable 5.1 Max EffortAnthropic83.41LiveBench 综合分推理91.7编程86.4智能体编程66.1推理91.7编程86.4智能体编程66.1详情
02Claude 5.5 Opus Thinking Max EffortAnthropic83.22LiveBench 综合分推理92.2编程89.3智能体编程71.7推理92.2编程89.3智能体编程71.7详情
03Claude Fable 5 Max EffortAnthropic82.97LiveBench 综合分推理89.7编程86.0智能体编程62.2推理89.7编程86.0智能体编程62.2详情
04GPT-6 Astra Max EffortOpenAI82.16LiveBench 综合分推理92.7编程80.4智能体编程57.3推理92.7编程80.4智能体编程57.3详情
05Claude 5.5 Opus Thinking xHigh EffortAnthropic82.06LiveBench 综合分推理90.7编程89.3智能体编程65.4推理90.7编程89.3智能体编程65.4详情
06GPT-6.1 Sol Max EffortOpenAI81.62LiveBench 综合分推理92.6编程80.4智能体编程54.5推理92.6编程80.4智能体编程54.5来源
07Muse Spark 1.3 xHigh EffortMeta81.59LiveBench 综合分推理89.7编程81.1智能体编程64.1推理89.7编程81.1智能体编程64.1详情
08DeepSeek V4.1 Flash Max EffortDeepSeek81.11LiveBench 综合分推理86.7编程80.0智能体编程77.3推理86.7编程80.0智能体编程77.3详情
09GPT-6.1 Sol xHigh EffortOpenAI81.11LiveBench 综合分推理91.6编程80.7智能体编程56.8推理91.6编程80.7智能体编程56.8来源
10GPT-5.6 Sol Max EffortOpenAI81.05LiveBench 综合分推理91.7编程83.9智能体编程56.2推理91.7编程83.9智能体编程56.2详情
11GPT-5.5 Thinking xHigh EffortOpenAI80.19LiveBench 综合分推理89.7编程82.1智能体编程54.0推理89.7编程82.1智能体编程54.0详情
12Claude 5 Opus Thinking Max EffortAnthropic80.08LiveBench 综合分推理91.2编程81.4智能体编程65.2推理91.2编程81.4智能体编程65.2详情
13Smaug AgenticAbacus.AI79.51LiveBench 综合分推理90.3编程82.5智能体编程64.6推理90.3编程82.5智能体编程64.6详情
14GPT-6 Sol Max EffortOpenAI79.25LiveBench 综合分推理88.7编程81.8智能体编程52.9推理88.7编程81.8智能体编程52.9详情
15Kimi K3Moonshot AI79.19LiveBench 综合分推理90.7编程81.4智能体编程62.2推理90.7编程81.4智能体编程62.2详情
16Gemini 3.7 Flash HighGoogle78.83LiveBench 综合分推理87.8编程78.9智能体编程58.3推理87.8编程78.9智能体编程58.3详情
17Qwen 3.8 MaxAlibaba78.46LiveBench 综合分推理88.2编程72.9智能体编程64.6推理88.2编程72.9智能体编程64.6详情
18Grok 4.6 xHighxAI78.04LiveBench 综合分推理90.5编程76.8智能体编程57.0推理90.5编程76.8智能体编程57.0来源
19GPT-5.4 Thinking xHigh EffortOpenAI77.97LiveBench 综合分推理88.1编程77.5智能体编程53.8推理88.1编程77.5智能体编程53.8详情
20Muse Spark 1.2 xHigh EffortMeta77.95LiveBench 综合分推理90.0编程77.5智能体编程57.6推理90.0编程77.5智能体编程57.6详情
21GPT-5.6 Terra Max EffortOpenAI77.94LiveBench 综合分推理90.6编程78.2智能体编程54.9推理90.6编程78.2智能体编程54.9详情
22Claude Sonnet 5.5 xHigh EffortAnthropic77.75LiveBench 综合分推理86.8编程88.9智能体编程39.3推理86.8编程88.9智能体编程39.3来源
23DeepSeek V4 Pro 0813DeepSeek77.44LiveBench 综合分推理85.8编程77.2智能体编程54.9推理85.8编程77.2智能体编程54.9详情
24Smaug FlashAbacus.AI77.44LiveBench 综合分推理86.2编程77.1智能体编程61.1推理86.2编程77.1智能体编程61.1详情
25Grok 4.7 xHighxAI77.40LiveBench 综合分推理82.7编程77.2智能体编程54.0推理82.7编程77.2智能体编程54.0详情
26Gemini 3.1 Pro Preview HighGoogle76.95LiveBench 综合分推理84.0编程76.5智能体编程44.1推理84.0编程76.5智能体编程44.1详情
27Smaug MiniAbacus.AI76.93LiveBench 综合分推理82.8编程75.4智能体编程60.8推理82.8编程75.4智能体编程60.8详情
28DeepSeek V4 Flash Vision ExpDeepSeek76.76LiveBench 综合分推理85.4编程68.2智能体编程65.1推理85.4编程68.2智能体编程65.1详情
29Claude 4.7 Opus Thinking xHigh EffortAnthropic76.53LiveBench 综合分推理87.2编程82.1智能体编程50.7推理87.2编程82.1智能体编程50.7详情
30Claude 4.8 Opus Thinking Max EffortAnthropic76.22LiveBench 综合分推理89.2编程81.8智能体编程50.5推理89.2编程81.8智能体编程50.5详情
31Qwen 3.8 Flash NextAlibaba76.19LiveBench 综合分推理87.4编程72.6智能体编程61.6推理87.4编程72.6智能体编程61.6详情
32GLM-5.3Z.ai76.14LiveBench 综合分推理85.8编程79.0智能体编程60.9推理85.8编程79.0智能体编程60.9详情
33Claude Sonnet 5 xHigh EffortAnthropic76.04LiveBench 综合分推理88.7编程80.7智能体编程59.4推理88.7编程80.7智能体编程59.4详情
34Gemini 3.8 Flash HighGoogle75.83LiveBench 综合分推理89.3编程72.5智能体编程54.2推理89.3编程72.5智能体编程54.2详情
35Grok 4.5xAI75.77LiveBench 综合分推理87.2编程68.6智能体编程56.5推理87.2编程68.6智能体编程56.5详情
36Claude Sonnet 5.5 Max EffortAnthropic75.67LiveBench 综合分推理91.6编程91.4智能体编程56.3推理91.6编程91.4智能体编程56.3来源
37Muse Spark 1.1 xHigh EffortMeta75.30LiveBench 综合分推理87.7编程77.2智能体编程58.5推理87.7编程77.2智能体编程58.5详情
38Qwen3.8 27BAlibaba75.27LiveBench 综合分推理80.0编程75.7智能体编程61.4推理80.0编程75.7智能体编程61.4详情
39Gemini 3.5 Flash HighGoogle74.64LiveBench 综合分推理82.0编程78.2智能体编程49.0推理82.0编程78.2智能体编程49.0详情
40GPT-5.2 HighOpenAI74.63LiveBench 综合分推理83.2编程76.1智能体编程50.3推理83.2编程76.1智能体编程50.3详情
41Claude 4.6 Opus Thinking High EffortAnthropic74.52LiveBench 综合分推理88.7编程78.2智能体编程49.0推理88.7编程78.2智能体编程49.0详情
42DeepSeek V4 Flash 0731DeepSeek74.17LiveBench 综合分推理86.6编程75.0智能体编程46.8推理86.6编程75.0智能体编程46.8详情
43GPT-5.2 CodexOpenAI73.98LiveBench 综合分推理77.7编程83.6智能体编程49.4推理77.7编程83.6智能体编程49.4详情
44Gemini 3.6 Flash HighGoogle73.59LiveBench 综合分推理85.1编程77.9智能体编程43.4推理85.1编程77.9智能体编程43.4详情
45GPT-5.6 Luna Max EffortOpenAI73.56LiveBench 综合分推理85.6编程82.9智能体编程48.4推理85.6编程82.9智能体编程48.4详情
46GLM-5.2Z.ai73.16LiveBench 综合分推理78.6编程79.7智能体编程51.8推理78.6编程79.7智能体编程51.8详情
47Qwen 3.7 MaxAlibaba73.14LiveBench 综合分推理83.3编程74.2智能体编程43.6推理83.3编程74.2智能体编程43.6详情
48Claude 4.6 Sonnet Thinking Medium EffortAnthropic72.99LiveBench 综合分推理84.8编程79.3智能体编程42.6推理84.8编程79.3智能体编程42.6详情
49Claude 4.5 Opus Thinking High EffortAnthropic72.58LiveBench 综合分推理80.1编程79.7智能体编程39.7推理80.1编程79.7智能体编程39.7详情
50GPT-6 Luna Max EffortOpenAI72.03LiveBench 综合分推理81.8编程79.0智能体编程51.2推理81.8编程79.0智能体编程51.2详情
显示更多已显示 50 / 64

题集版本 2026-06-25,仅在同一版本内比较。

未公布的数据记为 —