Skip to content

Leaderboard

Data source: AIHOT

Model data is shown as published (Chinese).

终端任务、应用开发、漏洞修补、改真实仓库、系统集成与线上排障、写代码、科学计算、整套程序的迁移与重建,看模型能不能把软件做出来。AIHOT 评分把各家难度不同的评测折算成同一种分数:50 分是参照模型的平均水平,每 15 分约是一个标准差;两个模型相差 3 分以上,先后基本不会变。评分不是正确率。分类评分只看这个领域,只列在这个领域测得够多的模型。

16 benchmarks · 10 organizations · Updated 2026-11-18 14:08 Beijing time

Domestic is inferred from vendor; open weights uses the model page’s weights link. Unranked models have no weights flag and are hidden by that filter. Prices: ¥ / million tokens.

RankModelReleasedBenchmarksCacheInputOutputScore
1Claude Fable 5.2Anthropic2026-11-0911 benchmarks¥1.33¥53.33¥266.6479.7
2Gemini 4 CarbonGoogle2026-10-287 benchmarks¥1.33¥26.66¥133.3275.2
3Claude Opus 5.5Anthropic2026-09-2216 benchmarks¥1.33¥26.66¥133.3274.6
4gpt-6.2-astraOpenAI2026-11-177 benchmarks¥2.67¥53.33¥266.6474.6
5gpt-6.2-solOpenAI2026-11-176 benchmarks¥0.67¥13.33¥66.6673.3
6Claude Sonnet 5.5Anthropic2026-09-2816 benchmarks¥0.67¥13.33¥66.6673.1
7Gemini 4 HydrogenGoogle2026-11-127 benchmarks¥0.5¥10¥5072.1
8Claude Fable 5.1Anthropic2026-09-0115 benchmarks¥1.67¥66.66¥333.3171.8
9GPT-6 AstraOpenAI2026-09-0316 benchmarks¥6.67¥66.66¥333.3170.4
10GPT-6.1 TerraOpenAI2026-10-269 benchmarks¥0.33¥6.67¥33.3370.3
11Gemini 4 ArgonGoogle2026-09-3015 benchmarks¥1.33¥26.66¥133.3269.1
12Qwen4 MaxAlibaba2026-11-164 benchmarks¥0.5¥10¥3068.3
13GPT-6.1 SolOpenAI2026-09-2915 benchmarks¥0.67¥13.33¥66.6668.0
14Claude Fable 5Anthropic2026-06-0911 benchmarks¥6.67¥66.66¥333.3168.0
15Claude Opus 5Anthropic2026-07-2414 benchmarks¥3.33¥33.33¥166.6567.9
16Hy4Tencent2026-11-046 benchmarks¥0.3¥6¥1866.7
17Muse EmberMeta2026-11-096 benchmarks¥0.0267¥0.67¥3.3365.8
18Kimi K3.1 ProMoonshot AI2026-10-239 benchmarks¥1.3¥13¥6865.7
19Muse WonderMeta2026-10-1212 benchmarks¥1.33¥26.66¥133.3265.6
20GPT-6 SolOpenAI2026-09-2214 benchmarks¥1.33¥13.33¥66.6665.2
21GPT-5.6 SolOpenAI2026-07-0914 benchmarks¥3.33¥33.33¥199.9864.3
22DeepSeek V4.1 Flash 1017DeepSeek2026-10-1711 benchmarks¥0.02¥1¥464.3
23Claude Haiku 5.5Anthropic2026-10-0716 benchmarks¥0.0667¥0.67¥3.3364.3
24MiMo-V2.6-ProXiaomi2026-09-2111 benchmarks¥0.025¥3¥663.5
25GLM-5.3Z.ai2026-08-1814 benchmarks¥2¥8¥2862.9
26Grok 4.7xAI2026-09-2116 benchmarks¥3.33¥13.33¥4062.5
27DeepSeek V4.1 FlashDeepSeek2026-09-1012 benchmarks¥0.02¥1¥462.3
28Qwen3.8 Max 0902Alibaba2026-09-0110 benchmarks¥1.5¥12¥3661.5
29Muse Spark 1.3Meta2026-09-0213 benchmarks¥1¥8.33¥28.3361.2
30Claude Opus 4.8Anthropic2026-05-2810 benchmarks¥3.33¥33.33¥166.6561.1
31Claude Opus 4.7Anthropic2026-04-165 benchmarks¥3.33¥33.33¥166.6560.4
32Hy 4 PreviewTencent2026-08-287 benchmarks¥0.3¥6¥1860.3
33GPT-5.6 TerraOpenAI2026-07-0912 benchmarks¥1.33¥13.33¥79.9960.1
34Step 5 PreviewStepFun2026-09-186 benchmarks¥0.7¥7¥2060.1
35Kimi K3Moonshot AI2026-07-1616 benchmarks¥2¥20¥10060.0
36Grok 4.6xAI2026-08-1212 benchmarks¥3.33¥13.33¥4060.0
37Inkling 2 SmallThinking Machines Lab2026-11-059 benchmarks¥0.2¥2¥6.6760.0
38Gemini 3.7 FlashGoogle2026-08-1311 benchmarks¥1¥10¥5060.0
39MiniMax-M3.1-FlashMiniMax2026-10-195 benchmarks¥0.1¥1¥459.8
40Claude Sonnet 5Anthropic2026-06-3011 benchmarks¥1.33¥13.33¥66.6659.7
41GPT-5.5OpenAI2026-04-238 benchmarks¥3.33¥33.33¥199.9859.3
42Gemini 3.8 FlashGoogle2026-09-0215 benchmarks¥1¥10¥5059.3
43DeepSeek V4 Pro 0813DeepSeek2026-08-1311 benchmarks¥0.15¥4.5¥13.559.2
44MiMo V 2.6 FlashXiaomi2026-09-219 benchmarks¥0.02¥1¥259.1
45GPT-6 LunaOpenAI2026-09-2215 benchmarks¥0.0667¥0.67¥3.3359.0
46Qwen3.8 MaxAlibaba2026-07-1910 benchmarks¥1.5¥12¥3658.9
47GLM-5.3 FlashZ.ai2026-08-2614 benchmarks¥0.23¥0.8¥2.858.7
48Qwen3.8 Flash NextAlibaba2026-08-264 benchmarks¥0.1¥0.8¥2.758.4
49Muse Spark 1.2Meta2026-08-059 benchmarks¥1¥8.33¥28.3358.2
50Muse Spark 1.1Meta2026-07-096 benchmarks¥1¥8.33¥28.3358.1

Blended usage: cache hits:input:output = 7:2:1. 2026-11-17 · $1 = ¥6.6661

Unranked · 3

Muse Wonder FullMeta

本轮未收录官方 API 价格或公开权重

68.3
ember-1Fireworks AI

评测数量还不够

62.9
Ling 3.1 Flash 1019InclusionAI

评测数量还不够

60.9
All models · 45
RankModelReleasedBenchmarksCacheInputOutputScore
51Grok 4.5xAI2026-07-0811 benchmarks¥2¥13.33¥4057.5
52DeepSeek V4 Flash 0731DeepSeek2026-07-317 benchmarks¥0.05¥1.5¥4.557.4
53GPT-5.4OpenAI2026-03-052 benchmarks¥1.67¥16.67¥99.9957.3
54Claude Opus 4.6Anthropic2026-02-054 benchmarks¥3.33¥33.33¥166.6557.2
55Mistral Large 4Mistral AI2026-10-068 benchmarks¥0.47¥4.67¥1456.7
56GPT-5.6 LunaOpenAI2026-07-0913 benchmarks¥0.13¥1.33¥856.5
57Qwen3.8-27BAlibaba2026-08-148 benchmarks¥0.6¥3¥1256.1
58Gemini 3.6 FlashGoogle2026-07-218 benchmarks¥1¥10¥5056.0
59GLM-5.2Z.ai2026-06-167 benchmarks¥2¥8¥2854.9
60Gemini 3.5 FlashGoogle2026-05-197 benchmarks¥1¥10¥59.9954.8
61Muse Spark OpenMeta2026-10-157 benchmarks———54.7
62DeepSeek V4 Pro PreviewDeepSeek2026-04-247 benchmarks¥0.025¥3¥652.7
63Claude Sonnet 4.6Anthropic2026-02-176 benchmarks¥2¥20¥99.9952.6
64Gemini 3.1 Pro PreviewGoogle2026-02-199 benchmarks¥1.33¥13.33¥79.9952.4
65Qwen3.7 MaxAlibaba2026-05-194 benchmarks¥2.4¥12¥3651.9
66Kimi K2.6Moonshot AI2026-04-205 benchmarks¥1.1¥6.5¥2751.7
67Kimi K2.7 CodeMoonshot AI2026-06-126 benchmarks¥1.3¥6.5¥2751.4
68Beam 1.0Reflection2026-10-169 benchmarks———50.5
69Qwen3.6 PlusAlibaba2026-04-013 benchmarks¥0.2¥2¥1250.0
70MiniMax-M3MiniMax2026-06-0111 benchmarks¥0.42¥2.1¥8.449.7
71GPT-5.4 Mini 2026-03-17OpenAI2026-03-177 benchmarks¥0.5¥5¥3049.1
72Gemini 3.5 Flash LiteGoogle2026-07-218 benchmarks¥0.2¥2¥16.6748.0
73Qwen3.7 PlusAlibaba2026-06-013 benchmarks¥0.4¥2¥847.7
74GLM-5.1Z.ai2026-04-075 benchmarks¥1.3¥6¥2447.7
75MiMo-V2.5-ProXiaomi2026-04-224 benchmarks¥0.025¥3¥647.5
76DeepSeek V4 Flash PreviewDeepSeek2026-04-243 benchmarks¥0.02¥1¥246.8
77GPT-5.4 Nano 2026-03-17OpenAI2026-03-174 benchmarks¥0.13¥1.33¥8.3346.5
78Nemotron 3.5 Super VLNVIDIA2026-10-155 benchmarks———44.9
79MiMo-V2.5Xiaomi2026-04-223 benchmarks¥0.02¥1¥244.8
80InklingThinking Machines Lab2026-07-1513 benchmarks¥1.13¥6.67¥2744.6
81Qwen3.6-27BAlibaba2026-04-224 benchmarks—¥3¥1844.3
82Grok 4.3xAI2026-04-175 benchmarks¥1.33¥8.33¥16.6743.4
83Inkling-SmallThinking Machines Lab2026-07-3010 benchmarks¥0.4¥2¥842.7
84Nemotron 3 UltraNVIDIA2026-06-046 benchmarks———42.2
85MiniMax M 2.7MiniMax2026-03-185 benchmarks¥0.42¥2.1¥8.441.5
86Muse GlimmerMeta2026-08-103 benchmarks———40.1
87Gemma 4 31 BGoogle2026-04-023 benchmarksFreeFreeFree39.1
88GPT-5 MiniOpenAI2025-08-073 benchmarks¥0.17¥1.67¥13.3336.9
89Claude Haiku 4.5 20251001Anthropic2025-10-153 benchmarks¥0.67¥6.67¥33.3335.7
90Gemini 2.5 ProGoogle2025-06-173 benchmarks¥0.83¥8.33¥66.6634.6
91Gemini 3.1 Flash Lite PreviewGoogle2026-03-035 benchmarks¥0.17¥1.67¥1032.4
92Mistral Medium 3.5Mistral AI2026-04-294 benchmarks¥1¥10¥5032.0
93Granite 4.2 8 BIBM2026-08-253 benchmarks———23.6
94gpt-oss-120bOpenAI2025-08-054 benchmarks———21.4
95Granite 4.2 3 BIBM2026-08-253 benchmarks———13.1