BenchLeader

全部模型

共有 1275 个模型拥有第三方基准测试结果。可在下方按名称或提供商搜索,也可以从使用最多的模型看起。任选两个模型对比。

735 / 735
#
1新Claude Fable 5.2max72.7±4$16.00109232 s1M2026-11-092026-11-11
2Gemini 4 Carbonhigh68.9±2$4.006473 s1.0M2026-10-282026-11-11
3Claude Opus 5.5max68.4±4$8.00116380 s1M2026-09-222026-11-11
4GPT-6 Astramax67.3±3$20.0041198 s1.1M2026-09-022026-11-11
5Claude Fable 5.1max66.7±3$20.0089151 s1M2026-09-012026-11-11
6GPT-6.1 Solmax66.6±2$4.0050167 s1.1M2026-09-292026-11-11
7Claude Fable 565.4±5$20.00503.21 s1M2026-06-092026-10-24
8GPT-5.6 Solmax65.3±3$8.0010034 s1.1M2026-07-092026-10-29
9Gemini 4 Argonhigh65.3±3$4.006646 s1.0M2026-09-302026-11-11
10Claude Opus 5max65.0±2$10.008049 s1M2026-07-242026-11-11
11Muse Wonder64.4±3$8.00703.29 s1.0M2026-10-122026-10-24
12Muse Spark(Apr 2026)64.2±4–––262k2026-04-082026-11-11
13Claude Sonnet 5.5max64.2±3$4.00152291 s1M2026-09-282026-11-11
14GPT-5.5xhigh64.1±4$11.258138 s1.1M2026-04-232026-11-11
15GPT-6.1 Terramax64.1±2$2.0072151 s1.1M2026-10-262026-11-11
16新Gemini 4 Hydrogenhigh63.4±2$3.0017016 s1.0M2026-11-122026-11-12
17GPT-5.4xhigh63.1±3$5.6373145 s1.1M2026-03-052026-10-13
18Claude Opus 4.763.1±4$10.00481.02 s1M2026-04-162026-11-11
19GPT-5.4 Pro63.1±4$67.5077.93 s1.1M2026-03-042026-10-07
20GPT-5.3-Codexxhigh62.9±7$4.819438 s400k2026-02-052026-10-29
21GPT-6 Solmax62.6±3$4.0012737 s1.1M2026-09-222026-11-11
22Gemini 3.5 Flashhigh62.5±3$3.381769.87 s1.0M2026-05-192026-11-11
23Claude Opus 4.6high62.5±3$10.00351.12 s1M2026-02-052026-11-11
24Muse Spark 1.162.2±6$2.001931.80 s1.0M2026-07-092026-11-11
25Gemini 3.7 Flashhigh62.1±4$1.50952.23 s1.0M2026-08-132026-11-11
26Qwen3.6 Maxmax62.1±3$2.92––262k2026-04-202026-11-11
27Kimi K3.1 Pro61.8±3$4.0030157 s1.0M2026-10-232026-11-13
28Muse Wonder Fullmax61.7±2–––1M2026-10-122026-11-11
29Muse Spark 1.361.6±8$2.001632.60 s1.0M2026-09-022026-10-13
30Muse Spark 1.2xhigh61.5±5$2.0020811 s1.0M2026-08-052026-11-11
31Claude Opus 4.8max61.1±4$10.008539 s1M2026-05-282026-10-13
32Gemini 3.6 Flashhigh61.0±3$1.5017711 s1.0M2026-07-212026-11-11
33Kimi K360.7±6$6.00760.85 s1.0M2026-07-162026-10-24
34Grok 4.6high60.6±3$3.007329 s500k2026-08-112026-11-11
35o360.4±3$3.501315.17 s200k2025-04-162026-11-11
36Gemini 3.1 Prohigh60.4±3$4.5011115 s1.0M2026-02-192026-10-29
37Gemini 3.8 Flashmedium60.2±5$1.5011615 s1.0M2026-09-022026-09-02
38GLM-5thinking60.2±4$1.556929 s205k2026-02-112026-10-13
39GPT-5.6 Terraxhigh60.1±4$4.509719 s1.1M2026-07-092026-11-11
40Gemini 3 Pro60.0±3––––2025-11-182026-11-11
41Qwen3.7max60.0±4$3.7517115 s1M2026-05-192026-11-11
42Kimi K3.1 Pro Togethermax60.0±2––––2026-10-232026-10-23
43GPT-5.2xhigh59.9±4$4.8176112 s400k2025-12-112026-10-13
44Claude Opus 4.5high59.7±3$10.00401.11 s200k2025-11-242026-11-11
45GPT-5 Pro59.6±6$41.25––400k2025-08-072025-11-13 · 数据陈旧
46Kimi K2.5thinking59.4±4$1.204342 s262k2026-01-272026-11-11
47Grok 4.559.3±4$3.00491.41 s500k2026-07-082026-11-11
48Kimi K2.6thinking59.3±7$1.714193 s262k2026-04-202026-04-20 · 数据陈旧
49Mimo v2 Pro59.1±5–––1M2026-03-182026-11-11
50Qwen3.6 Plus59.0±4$1.135244 s1M2026-03-312026-11-11
51GPT-5medium59.0±5$3.448223 s400k2025-08-072026-11-11
52Claude Sonnet 5high59.0±3$4.009916 s1M2026-06-302026-11-11
53GLM 5.2max58.9±4$2.156927 s1M2026-06-162026-11-11
54Gemini 3 Flashthinking58.9±8$1.131964.71 s1.0M2025-12-172025-12-17 · 数据陈旧
55GLM 5.3max58.8±3$2.158858 s1M2026-08-182026-11-11
56Mimo v2 Omni58.8±2–––256k2026-03-272026-11-11
57Qwen3.8 Maxmax58.7±3$3.0059120 s1M2026-08-032026-11-11
58GPT-5-Codexhigh58.6±5$3.441209.48 s272k2025-09-152026-09-01
59DeepSeek V4.1 Flashmax58.5±4$0.52524033 s1.0M2026-09-102026-11-11
60Grok 4.20 Multi-Agent58.5±2$1.5634912 s2M2026-03-092026-11-11
61Dola Seed 2.0 Pro58.4±1––––2026-02-142026-11-11
62Grok 4.20thinking58.3±4$1.5611112 s2M2026-02-172026-11-11
63GPT-5.3 Chat58.3±2$4.81––128k2026-03-032026-11-11
64GPT-5.6 Lunaxhigh58.2±4$0.45011028 s1.1M2026-07-092026-11-11
65Grok 458.1±3$6.006918 s256k2025-07-092026-11-11
66Claude Opus 4.1thinking58.1±3$30.00143.37 s200k2025-08-052026-11-11
67Muse Spark Open 816B A42Bxhigh58.1±6$2.00––1.0M2026-10-152026-10-15
68Inkling 2 Small57.9±3$0.47510049 s1.0M2026-11-052026-11-11
69Claude Sonnet 4.5high57.9±7$6.00400.91 s200k2025-09-292026-11-11
70GLM 5.3 Flash57.8±5$0.2385487 s1M2026-08-202026-11-11
71DeepSeek V4.1 Flash 1017max57.7±4$0.52517440 s1.0M2026-10-172026-11-11
72Qwen3.7 Plus57.6±7$0.7005750 s1M2026-06-022026-11-11
73Claude Sonnet 4.657.6±4$6.00430.59 s1M2026-02-172026-11-11
74GPT-5.2-Codex57.6±10$4.81451.46 s272k2025-12-182026-11-11
75GPT-5.1high57.5±4$3.448837 s400k2025-11-122026-11-11
76MiMo-V2.6-Pro57.5±4$0.5485442 s1.0M2026-09-212026-11-11
77Grok 4.7xhigh57.4±3$3.006856 s500k2026-09-212026-11-11
78Qwen3.5 Plusthinking57.4±1$0.900800.82 s1M2026-02-162026-10-13
79DeepSeek V4 Flash(Jul 2026)high57.2±4$0.262820.71 s1.0M2026-07-312026-11-11
80Grok 4.157.1±3––––2025-11-172026-11-11
81GPT-5.5 Instant56.9±6$11.251251.83 s400k2026-05-052026-11-11
82DeepSeek V3.2thinking56.7±4$0.3154430 s131k2025-12-012026-11-11
83Hy456.7±4$1.2594110 s960k2026-08-282026-11-11
84Claude Haiku 5.5xhigh56.7±3$0.20023150 s1M2026-10-072026-10-07
85MiniMax M2.756.6±5$0.5254219 s205k2026-03-182026-11-11
86GPT-5 mini56.6±8$0.688703.54 s400k2025-08-072026-02-17 · 数据陈旧
87Grok 4.3high56.5±7$1.568518 s1M2026-04-172026-10-01
88Qwen3.5 397B A17Bthinking56.4±9$1.355436 s262k2026-02-132026-03-02 · 数据陈旧
89MiMo-V2.556.3±4$0.138282.50 s1M2026-04-232026-11-11
90Step 3.5 Flash56.3±3$0.15013325 s256k2026-04-022026-11-11
91GPT-5 nano56.2±9$0.1381062.93 s400k2025-08-072025-11-24 · 数据陈旧
92MiniMax M356.1±4$0.52511860 s1M2026-06-012026-11-11
93Gemma 4 26B A4B56.1±8$0.262760.41 s262k2026-04-022026-11-11
94Step 5 Preview56.0±4$1.4312424 s1M2026-09-182026-11-11
95DeepSeek V4 Pro(Aug 2026)high56.0±4$1.988043 s1.0M2026-08-132026-11-11
96o3-prohigh55.9±9$35.00––200k2025-06-102026-10-07
97LongCat Flash55.9±7––––2025-09-012026-11-11
98Qwen3.6 27Bthinking55.9±7$1.356049 s262k2026-04-222026-04-22 · 数据陈旧
99Claude Opus 4thinking55.7±4$30.00––200k2025-05-222026-11-11
100Hy355.7±7$0.2437744 s262k2026-07-062026-11-11

仅见于价格表

有 443 个模型只出现在价格表中,未列入表格。以下是大家常搜索的几个;它们的页面显示标价,暂无第三方基准测试结果。

已下架

我们读取的基准测试发布方都已不再收录的模型。每个页面保留我们最后获得的数据,以及这些数据最后有效的日期。