跳到正文

模型榜

数据来源:AIHOT

事实准确、长文理解、语言与按要求作答、识破错误前提,看模型知道多少、答得多稳。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。

8 项评测 · 6 家机构 · 更新 2026-11-18 20:05 北京时间

国产由厂商推断,开源依据模型页的权重链接;暂不排名的模型缺少权重标记,开源筛选时不展示。价格单位:¥ / 百万 Token。

名次模型上线日期评测缓存价输入价输出价评分
1Claude Fable 5.2Anthropic2026-11-096 项评测¥1.33¥53.33¥266.6482.3
2gpt-6.2-astraOpenAI2026-11-175 项评测¥2.67¥53.33¥266.6479.7
3Gemini 4 CarbonGoogle2026-10-286 项评测¥1.33¥26.66¥133.3279.4
4Gemini 4 HydrogenGoogle2026-11-123 项评测¥0.5¥10¥5076.4
5gpt-6.2-solOpenAI2026-11-174 项评测¥0.67¥13.33¥66.6674.7
6GPT-6 AstraOpenAI2026-09-038 项评测¥6.67¥66.66¥333.3174.6
7Gemini 4 ArgonGoogle2026-09-307 项评测¥1.33¥26.66¥133.3273.3
8GPT-6.1 SolOpenAI2026-09-298 项评测¥0.67¥13.33¥66.6672.3
9Claude Fable 5.1Anthropic2026-09-017 项评测¥1.67¥66.66¥333.3170.9
10Claude Opus 5.5Anthropic2026-09-228 项评测¥1.33¥26.66¥133.3270.5
11Kimi K3.1 ProMoonshot AI2026-10-236 项评测¥1.3¥13¥6870.3
12Claude Fable 5Anthropic2026-06-096 项评测¥6.67¥66.66¥333.3169.8
13Qwen4 MaxAlibaba2026-11-165 项评测¥0.5¥10¥3069.5
14Claude Opus 5Anthropic2026-07-247 项评测¥3.33¥33.33¥166.6569.0
15Gemini 3.8 FlashGoogle2026-09-027 项评测¥1¥10¥5068.6
16GPT-5.6 SolOpenAI2026-07-096 项评测¥3.33¥33.33¥199.9867.8
17Kimi K3Moonshot AI2026-07-168 项评测¥2¥20¥10066.8
18GPT-6.1 TerraOpenAI2026-10-266 项评测¥0.33¥6.67¥33.3366.6
19Hy4Tencent2026-11-045 项评测¥0.3¥6¥1866.4
20Gemini 3.7 FlashGoogle2026-08-136 项评测¥1¥10¥5066.2
21DeepSeek V4.1 Flash 1017DeepSeek2026-10-176 项评测¥0.02¥1¥465.9
22Muse WonderMeta2026-10-126 项评测¥1.33¥26.66¥133.3265.5
23GPT-5.5OpenAI2026-04-236 项评测¥3.33¥33.33¥199.9865.3
24GPT-6 SolOpenAI2026-09-227 项评测¥1.33¥13.33¥66.6665.0
25Grok 4.7xAI2026-09-218 项评测¥3.33¥13.33¥4065.0
26Claude Sonnet 5.5Anthropic2026-09-288 项评测¥0.67¥13.33¥66.6664.6
27Muse Spark 1.3Meta2026-09-025 项评测¥1¥8.33¥28.3364.6
28Claude Opus 4.8Anthropic2026-05-286 项评测¥3.33¥33.33¥166.6564.1
29Muse EmberMeta2026-11-093 项评测¥0.0267¥0.67¥3.3362.7
30Gemini 3.1 Pro PreviewGoogle2026-02-196 项评测¥1.33¥13.33¥79.9962.4
31Grok 4.6xAI2026-08-127 项评测¥3.33¥13.33¥4062.3
32DeepSeek V4.1 FlashDeepSeek2026-09-106 项评测¥0.02¥1¥462.0
33Qwen3.8 MaxAlibaba2026-07-197 项评测¥1.5¥12¥3661.9
34Inkling 2 SmallThinking Machines Lab2026-11-053 项评测¥0.2¥2¥6.6761.8
35GLM-5.3Z.ai2026-08-188 项评测¥2¥8¥2861.5
36Gemini 3.6 FlashGoogle2026-07-216 项评测¥1¥10¥5060.8
37Gemini 3.5 FlashGoogle2026-05-196 项评测¥1¥10¥59.9960.7
38Muse Spark 1.2Meta2026-08-056 项评测¥1¥8.33¥28.3360.6
39Qwen3.7 MaxAlibaba2026-05-196 项评测¥2.4¥12¥3660.6
40Step 5 PreviewStepFun2026-09-185 项评测¥0.7¥7¥2060.5
41Qwen3.8 Max 0902Alibaba2026-09-014 项评测¥1.5¥12¥3660.5
42Grok 4.5xAI2026-07-086 项评测¥2¥13.33¥4060.1
43MiMo-V2.6-ProXiaomi2026-09-214 项评测¥0.025¥3¥660.1
44GPT-5.6 TerraOpenAI2026-07-097 项评测¥1.33¥13.33¥79.9959.7
45Hy 4 PreviewTencent2026-08-282 项评测¥0.3¥6¥1859.5
46Claude Opus 4.6Anthropic2026-02-055 项评测¥3.33¥33.33¥166.6559.4
47Claude Opus 4.7Anthropic2026-04-165 项评测¥3.33¥33.33¥166.6559.0
48GPT-5.4OpenAI2026-03-054 项评测¥1.67¥16.67¥99.9958.9
49Muse Spark 1.1Meta2026-07-095 项评测¥1¥8.33¥28.3358.7
50Muse Spark OpenMeta2026-10-154 项评测———58.3

一般用量:缓存命中:输入:输出 = 7:2:1。 2026-11-17 · $1 = ¥6.6661

暂不排名 · 2

Muse Wonder FullMeta

本轮未收录官方 API 价格或公开权重

68.4
Qwen3.8 2.4 T A 95 BAlibaba

评测数量还不够

59.6
全部模型 · 45
名次模型上线日期评测缓存价输入价输出价评分
51Qwen3.8 Flash NextAlibaba2026-08-264 项评测¥0.1¥0.8¥2.758.0
52Claude Haiku 5.5Anthropic2026-10-077 项评测¥0.0667¥0.67¥3.3356.8
53DeepSeek V4 Pro 0813DeepSeek2026-08-136 项评测¥0.15¥4.5¥13.556.7
54GPT-6 LunaOpenAI2026-09-227 项评测¥0.0667¥0.67¥3.3355.9
55Claude Sonnet 5Anthropic2026-06-306 项评测¥1.33¥13.33¥66.6655.3
56GLM-5.3 FlashZ.ai2026-08-265 项评测¥0.23¥0.8¥2.854.0
57DeepSeek V4 Flash 0731DeepSeek2026-07-315 项评测¥0.05¥1.5¥4.554.0
58Beam 1.0Reflection2026-10-163 项评测———53.7
59Claude Sonnet 4.6Anthropic2026-02-176 项评测¥2¥20¥99.9953.5
60Qwen3.8-27BAlibaba2026-08-146 项评测¥0.6¥3¥1253.4
61GPT-5.6 LunaOpenAI2026-07-097 项评测¥0.13¥1.33¥853.3
62Kimi K2.6Moonshot AI2026-04-206 项评测¥1.1¥6.5¥2753.2
63MiniMax-M3.1-FlashMiniMax2026-10-195 项评测¥0.1¥1¥452.9
64MiniMax-M3MiniMax2026-06-017 项评测¥0.42¥2.1¥8.452.6
65Gemini 3.5 Flash LiteGoogle2026-07-216 项评测¥0.2¥2¥16.6752.6
66Qwen3.6 PlusAlibaba2026-04-015 项评测¥0.2¥2¥1252.1
67MiMo V 2.6 FlashXiaomi2026-09-213 项评测¥0.02¥1¥252.0
68GLM-5.2Z.ai2026-06-167 项评测¥2¥8¥2851.1
69InklingThinking Machines Lab2026-07-155 项评测¥1.13¥6.67¥2750.9
70Mistral Large 4Mistral AI2026-10-062 项评测¥0.47¥4.67¥1450.6
71Kimi K2.7 CodeMoonshot AI2026-06-124 项评测¥1.3¥6.5¥2750.5
72DeepSeek V4 Pro PreviewDeepSeek2026-04-247 项评测¥0.025¥3¥649.7
73Nemotron 3 UltraNVIDIA2026-06-045 项评测———49.6
74Qwen3.7 PlusAlibaba2026-06-013 项评测¥0.4¥2¥849.6
75Grok 4.3xAI2026-04-176 项评测¥1.33¥8.33¥16.6747.6
76Nemotron 3.5 Super VLNVIDIA2026-10-153 项评测———47.3
77GLM-5.1Z.ai2026-04-074 项评测¥1.3¥6¥2447.1
78MiMo-V2.5-ProXiaomi2026-04-224 项评测¥0.025¥3¥646.8
79DeepSeek V4 Flash PreviewDeepSeek2026-04-243 项评测¥0.02¥1¥246.7
80GPT-5.4 Mini 2026-03-17OpenAI2026-03-176 项评测¥0.5¥5¥3046.5
81MiMo-V2.5Xiaomi2026-04-222 项评测¥0.02¥1¥244.4
82Qwen3.6-27BAlibaba2026-04-223 项评测—¥3¥1844.1
83Inkling-SmallThinking Machines Lab2026-07-304 项评测¥0.4¥2¥842.3
84GPT-5.4 Nano 2026-03-17OpenAI2026-03-175 项评测¥0.13¥1.33¥8.3341.2
85Muse GlimmerMeta2026-08-102 项评测———40.0
86GPT-5 MiniOpenAI2025-08-073 项评测¥0.17¥1.67¥13.3336.8
87Gemma 4 31 BGoogle2026-04-022 项评测免费免费免费35.5
88Gemini 2.5 ProGoogle2025-06-173 项评测¥0.83¥8.33¥66.6633.6
89MiniMax M 2.7MiniMax2026-03-184 项评测¥0.42¥2.1¥8.433.3
90Claude Haiku 4.5 20251001Anthropic2025-10-152 项评测¥0.67¥6.67¥33.3333.3
91Mistral Medium 3.5Mistral AI2026-04-294 项评测¥1¥10¥5030.2
92Gemini 3.1 Flash Lite PreviewGoogle2026-03-035 项评测¥0.17¥1.67¥1027.1
93Granite 4.2 8 BIBM2026-08-252 项评测———25.5
94gpt-oss-120bOpenAI2025-08-054 项评测———18.2
95Granite 4.2 3 BIBM2026-08-252 项评测———11.5