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。
| 排名 | 模型 / 开发机构 | LiveBench 综合分 | 推理 | 编程 | 智能体编程 | 详情 | |
|---|---|---|---|---|---|---|---|
| 01 | Claude Fable 5.1 Max EffortAnthropic | 83.41LiveBench 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 |
| 02 | Claude 5.5 Opus Thinking Max EffortAnthropic | 83.22LiveBench 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 |
| 03 | Claude Fable 5 Max EffortAnthropic | 82.97LiveBench 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 |
| 04 | GPT-6 Astra Max EffortOpenAI | 82.16LiveBench 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 |
| 05 | Claude 5.5 Opus Thinking xHigh EffortAnthropic | 82.06LiveBench 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 |
| 06 | GPT-6.1 Sol Max EffortOpenAI | 81.62LiveBench 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 |
| 07 | Muse Spark 1.3 xHigh EffortMeta | 81.59LiveBench 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 |
| 08 | DeepSeek V4.1 Flash Max EffortDeepSeek | 81.11LiveBench 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 |
| 09 | GPT-6.1 Sol xHigh EffortOpenAI | 81.11LiveBench 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 |
| 10 | GPT-5.6 Sol Max EffortOpenAI | 81.05LiveBench 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 |
| 11 | GPT-5.5 Thinking xHigh EffortOpenAI | 80.19LiveBench 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 |
| 12 | Claude 5 Opus Thinking Max EffortAnthropic | 80.08LiveBench 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 |
| 13 | Smaug AgenticAbacus.AI | 79.51LiveBench 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 |
| 14 | GPT-6 Sol Max EffortOpenAI | 79.25LiveBench 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 |
| 15 | Kimi K3Moonshot AI | 79.19LiveBench 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 |
| 16 | Gemini 3.7 Flash HighGoogle | 78.83LiveBench 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 |
| 17 | Qwen 3.8 MaxAlibaba | 78.46LiveBench 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 |
| 18 | Grok 4.6 xHighxAI | 78.04LiveBench 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 |
| 19 | GPT-5.4 Thinking xHigh EffortOpenAI | 77.97LiveBench 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 |
| 20 | Muse Spark 1.2 xHigh EffortMeta | 77.95LiveBench 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 |
| 21 | GPT-5.6 Terra Max EffortOpenAI | 77.94LiveBench 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 |
| 22 | Claude Sonnet 5.5 xHigh EffortAnthropic | 77.75LiveBench 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 |
| 23 | DeepSeek V4 Pro 0813DeepSeek | 77.44LiveBench 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 |
| 24 | Smaug FlashAbacus.AI | 77.44LiveBench 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 |
| 25 | Grok 4.7 xHighxAI | 77.40LiveBench 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 |
| 26 | Gemini 3.1 Pro Preview HighGoogle | 76.95LiveBench 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 |
| 27 | Smaug MiniAbacus.AI | 76.93LiveBench 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 |
| 28 | DeepSeek V4 Flash Vision ExpDeepSeek | 76.76LiveBench 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 |
| 29 | Claude 4.7 Opus Thinking xHigh EffortAnthropic | 76.53LiveBench 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 |
| 30 | Claude 4.8 Opus Thinking Max EffortAnthropic | 76.22LiveBench 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 |
| 31 | Qwen 3.8 Flash NextAlibaba | 76.19LiveBench 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 |
| 32 | GLM-5.3Z.ai | 76.14LiveBench 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 |
| 33 | Claude Sonnet 5 xHigh EffortAnthropic | 76.04LiveBench 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 |
| 34 | Gemini 3.8 Flash HighGoogle | 75.83LiveBench 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 |
| 35 | Grok 4.5xAI | 75.77LiveBench 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 |
| 36 | Claude Sonnet 5.5 Max EffortAnthropic | 75.67LiveBench 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 |
| 37 | Muse Spark 1.1 xHigh EffortMeta | 75.30LiveBench 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 |
| 38 | Qwen3.8 27BAlibaba | 75.27LiveBench 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 |
| 39 | Gemini 3.5 Flash HighGoogle | 74.64LiveBench 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 |
| 40 | GPT-5.2 HighOpenAI | 74.63LiveBench 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 |
| 41 | Claude 4.6 Opus Thinking High EffortAnthropic | 74.52LiveBench 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 |
| 42 | DeepSeek V4 Flash 0731DeepSeek | 74.17LiveBench 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 |
| 43 | GPT-5.2 CodexOpenAI | 73.98LiveBench 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 |
| 44 | Gemini 3.6 Flash HighGoogle | 73.59LiveBench 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 |
| 45 | GPT-5.6 Luna Max EffortOpenAI | 73.56LiveBench 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 |
| 46 | GLM-5.2Z.ai | 73.16LiveBench 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 |
| 47 | Qwen 3.7 MaxAlibaba | 73.14LiveBench 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 |
| 48 | Claude 4.6 Sonnet Thinking Medium EffortAnthropic | 72.99LiveBench 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 |
| 49 | Claude 4.5 Opus Thinking High EffortAnthropic | 72.58LiveBench 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 |
| 50 | GPT-6 Luna Max EffortOpenAI | 72.03LiveBench 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 |