跳到正文

模型榜

数据来源:AIHOT

只用第三方评测机构自己给模型实测的成绩,厂商自己公布的分数一概不计;各项成绩折算成同一种分数后按领域加权:编程 38%、科研 19%、专业办公 19%、知识问答 14%、真人盲选 10%。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。

54 项评测 · 16 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 5.2Anthropic2026-11-0940 项评测¥1.33¥53.33¥266.6479.9
2Gemini 4 CarbonGoogle2026-10-2831 项评测¥1.33¥26.66¥133.3277.1
3gpt-6.2-astraOpenAI2026-11-1730 项评测¥2.67¥53.33¥266.6476.4
4Claude Opus 5.5Anthropic2026-09-2257 项评测¥1.33¥26.66¥133.3274.1
5Gemini 4 HydrogenGoogle2026-11-1222 项评测¥0.5¥10¥5073.4
6gpt-6.2-solOpenAI2026-11-1726 项评测¥0.67¥13.33¥66.6673.0
7Claude Fable 5.1Anthropic2026-09-0150 项评测¥1.67¥66.66¥333.3172.3
8GPT-6 AstraOpenAI2026-09-0357 项评测¥6.67¥66.66¥333.3171.9
9Gemini 4 ArgonGoogle2026-09-3051 项评测¥1.33¥26.66¥133.3271.7
10Claude Sonnet 5.5Anthropic2026-09-2857 项评测¥0.67¥13.33¥66.6671.5
11GPT-6.1 SolOpenAI2026-09-2956 项评测¥0.67¥13.33¥66.6670.0
12Qwen4 MaxAlibaba2026-11-1624 项评测¥0.5¥10¥3069.3
13Claude Opus 5Anthropic2026-07-2447 项评测¥3.33¥33.33¥166.6569.2
14GPT-6.1 TerraOpenAI2026-10-2632 项评测¥0.33¥6.67¥33.3369.0
15Claude Fable 5Anthropic2026-06-0938 项评测¥6.67¥66.66¥333.3168.7
16Kimi K3.1 ProMoonshot AI2026-10-2338 项评测¥1.3¥13¥6867.8
17Muse WonderMeta2026-10-1243 项评测¥1.33¥26.66¥133.3267.3
18GPT-5.6 SolOpenAI2026-07-0943 项评测¥3.33¥33.33¥199.9866.7
19Hy4Tencent2026-11-0424 项评测¥0.3¥6¥1866.7
20Muse EmberMeta2026-11-0922 项评测¥0.0267¥0.67¥3.3366.0
21DeepSeek V4.1 Flash 1017DeepSeek2026-10-1739 项评测¥0.02¥1¥465.8
22GPT-6 SolOpenAI2026-09-2247 项评测¥1.33¥13.33¥66.6665.5
23Muse Spark 1.3Meta2026-09-0242 项评测¥1¥8.33¥28.3364.2
24Gemini 3.8 FlashGoogle2026-09-0255 项评测¥1¥10¥5063.6
25Grok 4.7xAI2026-09-2157 项评测¥3.33¥13.33¥4063.5
26GLM-5.3Z.ai2026-08-1853 项评测¥2¥8¥2863.1
27Kimi K3Moonshot AI2026-07-1655 项评测¥2¥20¥10062.9
28Qwen3.8 Max 0902Alibaba2026-09-0130 项评测¥1.5¥12¥3662.9
29GPT-5.5OpenAI2026-04-2333 项评测¥3.33¥33.33¥199.9862.8
30MiMo-V2.6-ProXiaomi2026-09-2136 项评测¥0.025¥3¥662.7
31Gemini 3.7 FlashGoogle2026-08-1338 项评测¥1¥10¥5062.5
32Claude Haiku 5.5Anthropic2026-10-0754 项评测¥0.0667¥0.67¥3.3362.3
33DeepSeek V4.1 FlashDeepSeek2026-09-1047 项评测¥0.02¥1¥462.3
34Grok 4.6xAI2026-08-1245 项评测¥3.33¥13.33¥4062.3
35Inkling 2 SmallThinking Machines Lab2026-11-0530 项评测¥0.2¥2¥6.6762.2
36Claude Opus 4.8Anthropic2026-05-2837 项评测¥3.33¥33.33¥166.6562.2
37Hy 4 PreviewTencent2026-08-2819 项评测¥0.3¥6¥1862.1
38Qwen3.8 MaxAlibaba2026-07-1942 项评测¥1.5¥12¥3662.1
39GPT-5.6 TerraOpenAI2026-07-0941 项评测¥1.33¥13.33¥79.9962.0
40Claude Opus 4.7Anthropic2026-04-1629 项评测¥3.33¥33.33¥166.6560.7
41Step 5 PreviewStepFun2026-09-1832 项评测¥0.7¥7¥2060.5
42Muse Spark 1.2Meta2026-08-0532 项评测¥1¥8.33¥28.3360.0
43MiniMax-M3.1-FlashMiniMax2026-10-1926 项评测¥0.1¥1¥459.8
44Qwen3.8 Flash NextAlibaba2026-08-2620 项评测¥0.1¥0.8¥2.759.6
45DeepSeek V4 Pro 0813DeepSeek2026-08-1346 项评测¥0.15¥4.5¥13.559.6
46Claude Sonnet 5Anthropic2026-06-3037 项评测¥1.33¥13.33¥66.6659.3
47MiMo V 2.6 FlashXiaomi2026-09-2130 项评测¥0.02¥1¥258.9
48Grok 4.5xAI2026-07-0836 项评测¥2¥13.33¥4058.9
49GPT-5.4OpenAI2026-03-0517 项评测¥1.67¥16.67¥99.9958.8
50GLM-5.3 FlashZ.ai2026-08-2649 项评测¥0.23¥0.8¥2.858.6

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 · 6

Muse Wonder FullMeta

本轮未收录官方 API 价格或公开权重

70.1
ember-1Fireworks AI

评测数量还不够

63.3
Ling 3.1 Flash 1019InclusionAI

评测数量还不够

61.1
Ling 3.1 FlashInclusionAI

评测数量还不够

60
Qwen3.8 2.4 T A 95 BAlibaba

评测数量还不够

60
hy4-flashTencent

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
51GPT-6 LunaOpenAI2026-09-2253 项评测¥0.0667¥0.67¥3.3358.5
52Claude Opus 4.6Anthropic2026-02-0523 项评测¥3.33¥33.33¥166.6558.4
53Muse Spark 1.1Meta2026-07-0925 项评测¥1¥8.33¥28.3358.3
54DeepSeek V4 Flash 0731DeepSeek2026-07-3131 项评测¥0.05¥1.5¥4.558.0
55Gemini 3.6 FlashGoogle2026-07-2135 项评测¥1¥10¥5057.2
56GPT-5.6 LunaOpenAI2026-07-0944 项评测¥0.13¥1.33¥857.1
57Muse Spark OpenMeta2026-10-1527 项评测———57.0
58Gemini 3.5 FlashGoogle2026-05-1934 项评测¥1¥10¥59.9956.7
59Mistral Large 4Mistral AI2026-10-0627 项评测¥0.47¥4.67¥1455.8
60Qwen3.8-27BAlibaba2026-08-1434 项评测¥0.6¥3¥1255.7
61Gemini 3.1 Pro PreviewGoogle2026-02-1939 项评测¥1.33¥13.33¥79.9955.4
62GLM-5.2Z.ai2026-06-1632 项评测¥2¥8¥2854.9
63Qwen3.7 MaxAlibaba2026-05-1927 项评测¥2.4¥12¥3654.3
64Claude Sonnet 4.6Anthropic2026-02-1730 项评测¥2¥20¥99.9953.5
65Beam 1.0Reflection2026-10-1630 项评测———52.6
66DeepSeek V4 Pro PreviewDeepSeek2026-04-2434 项评测¥0.025¥3¥652.5
67Kimi K2.6Moonshot AI2026-04-2028 项评测¥1.1¥6.5¥2751.2
68MiniMax-M3MiniMax2026-06-0144 项评测¥0.42¥2.1¥8.451.0
69Kimi K2.7 CodeMoonshot AI2026-06-1223 项评测¥1.3¥6.5¥2750.8
70Qwen3.6 PlusAlibaba2026-04-0123 项评测¥0.2¥2¥1248.5
71GLM-5.1Z.ai2026-04-0721 项评测¥1.3¥6¥2448.3
72Qwen3.7 PlusAlibaba2026-06-0116 项评测¥0.4¥2¥848.0
73DeepSeek V4 Flash PreviewDeepSeek2026-04-2418 项评测¥0.02¥1¥247.6
74MiMo-V2.5-ProXiaomi2026-04-2219 项评测¥0.025¥3¥647.5
75GPT-5.4 Mini 2026-03-17OpenAI2026-03-1731 项评测¥0.5¥5¥3047.2
76InklingThinking Machines Lab2026-07-1544 项评测¥1.13¥6.67¥2747.0
77Gemini 3.5 Flash LiteGoogle2026-07-2138 项评测¥0.2¥2¥16.6746.4
78Nemotron 3.5 Super VLNVIDIA2026-10-1518 项评测———45.8
79Qwen3.6-27BAlibaba2026-04-2218 项评测—¥3¥1844.1
80Inkling-SmallThinking Machines Lab2026-07-3034 项评测¥0.4¥2¥844.0
81Grok 4.3xAI2026-04-1728 项评测¥1.33¥8.33¥16.6743.8
82MiMo-V2.5Xiaomi2026-04-2215 项评测¥0.02¥1¥243.8
83Nemotron 3 UltraNVIDIA2026-06-0426 项评测———43.6
84GPT-5.4 Nano 2026-03-17OpenAI2026-03-1726 项评测¥0.13¥1.33¥8.3343.1
85Muse GlimmerMeta2026-08-1015 项评测———39.5
86MiniMax M 2.7MiniMax2026-03-1820 项评测¥0.42¥2.1¥8.439.0
87Gemma 4 31 BGoogle2026-04-0214 项评测免费免费免费37.2
88GPT-5 MiniOpenAI2025-08-0717 项评测¥0.17¥1.67¥13.3336.2
89Claude Haiku 4.5 20251001Anthropic2025-10-1512 项评测¥0.67¥6.67¥33.3333.6
90Gemini 2.5 ProGoogle2025-06-1716 项评测¥0.83¥8.33¥66.6633.2
91Mistral Medium 3.5Mistral AI2026-04-2921 项评测¥1¥10¥5031.6
92Gemini 3.1 Flash Lite PreviewGoogle2026-03-0323 项评测¥0.17¥1.67¥1030.4
93Granite 4.2 8 BIBM2026-08-2513 项评测———22.3
94gpt-oss-120bOpenAI2025-08-0518 项评测———20.7
95Granite 4.2 3 BIBM2026-08-2513 项评测———11.3