1
A
claude-opus-4-6
claude-opus-4-6Anthropic
100.0
3.1K
1M
¥36 / ¥180输入/输出
对话 · 文本 · 专家榜
基于公开偏好评测数据展示 文本 / 专家榜 榜单，每个榜单独立排序。
专家榜模型怎么选？
专家榜排行榜用于回答“专家榜哪个 AI 模型更强”这类选型问题。页面基于公开偏好评测数据展示模型排名，每个榜单独立排序，不把不同任务的原始分数直接混合。当前榜首是 claude-opus-4-6，厂商为 anthropic，百分制分数 100.0，样本量 3.1K。 数据来源保留 Arena 原始榜单和 LMArena leaderboard dataset，更新时间以榜单发布时间 2026-05-27 为准。
claude-opus-4-6claude-opus-4-6-thinkinggpt-5.5-highgpt-5.4-highclaude-opus-4-7
2
Aclaude-opus-4-6-thinking
claude-opus-4-6-thinkingAnthropic
99.7
2.6K
1M
¥36 / ¥180输入/输出
3
O
gpt-5.5-high
gpt-5.5-highOpenai
99.4
1.5K
1.05M
¥36 / ¥216输入/输出
4
Ogpt-5.4-high
gpt-5.4-highOpenai
99.0
2.4K
1.05M
¥18 / ¥108输入/输出
5
Aclaude-opus-4-7
claude-opus-4-7Anthropic
98.7
2K
1M
¥36 / ¥180输入/输出
6
MI
mimo-v2.5-pro
mimo-v2.5-proXiaomi
98.4
1.4K
1.05M
¥7.2 / ¥21.6输入/输出
7
G
gemini-3.5-flash
gemini-3.5-flashGoogle
98.1
899
1.05M
¥10.8 / ¥64.8输入/输出
8
Aclaude-opus-4-7-thinking
claude-opus-4-7-thinkingAnthropic
97.7
1.9K
1M
¥36 / ¥180输入/输出
9
A
amazon-nova-experimental-chat-26-02-10
amazon-nova-experimental-chat-26-02-10Amazon
97.4
270
-
-
10
Ogpt-5.5
gpt-5.5Openai
97.1
1.6K
1.05M
¥36 / ¥216输入/输出
11
Aclaude-sonnet-4-6
claude-sonnet-4-6Anthropic
96.8
2.5K
1M
¥21.6 / ¥108输入/输出
12
Ggemini-3.1-pro-preview
gemini-3.1-pro-previewGoogle
96.4
3.8K
1.05M
¥14.4 / ¥86.4输入/输出
13
A
qwen3.5-max-preview
qwen3.5-max-previewAlibaba
96.1
1.9K
-
-
14
Aqwen3.7-max-preview
qwen3.7-max-previewAlibaba
95.8
345
1M
¥18 / ¥54输入/输出
15
Ogpt-5.4
gpt-5.4Openai
95.5
2.5K
1.05M
¥18 / ¥108输入/输出
16
M
kimi-k2.6
kimi-k2.6Moonshot
95.1
1.3K
262K
¥6.84 / ¥28.8输入/输出
17
Z
glm-5.1
glm-5.1Zai
94.8
1.2K
200K
¥0 / ¥0输入/输出
18
Aclaude-opus-4-5-20251101
claude-opus-4-5-20251101Anthropic
94.5
4.6K
200K
¥36 / ¥180输入/输出
19
Aclaude-sonnet-4-5-20250929-thinking-32k
claude-sonnet-4-5-20250929-thinking-32kAnthropic
94.2
5K
200K
¥21.6 / ¥108输入/输出
20
Aclaude-opus-4-5-20251101-thinking-32k
claude-opus-4-5-20251101-thinking-32kAnthropic
93.8
2.2K
200K
¥108 / ¥540输入/输出
21
Ggemini-3-pro
gemini-3-proGoogle
93.5
2.5K
1.05M
¥14.4 / ¥86.4输入/输出
22
MImimo-v2-pro
mimo-v2-proXiaomi
93.2
1.9K
1.05M
¥7.2 / ¥21.6输入/输出
23
Aclaude-sonnet-4-5-20250929
claude-sonnet-4-5-20250929Anthropic
92.9
5.1K
200K
¥21.6 / ¥108输入/输出
24
Aqwen3.6-max-preview
qwen3.6-max-previewAlibaba
92.5
469
246K
¥9.5 / ¥56.9输入/输出
25
B
ernie-5.1
ernie-5.1Baidu
92.2
1.2K
119K
¥5.4 / ¥21.6输入/输出
26
Ogpt-5.1-high
gpt-5.1-highOpenai
91.9
2.3K
400K
¥9 / ¥72输入/输出
27
Aqwen3.5-397b-a17b
qwen3.5-397b-a17bAlibaba
91.6
2.8K
262K
¥3.1 / ¥18.6输入/输出
28
Mkimi-k2.5-thinking
kimi-k2.5-thinkingMoonshot
91.2
2.8K
262K
¥4.32 / ¥21.6输入/输出
29
Ggemini-3-flash
gemini-3-flashGoogle
90.9
1.9K
1.05M
¥3.6 / ¥21.6输入/输出
30
MImimo-v2.5
mimo-v2.5Xiaomi
90.6
1.5K
1.05M
¥2.88 / ¥14.4输入/输出
31
Zglm-5
glm-5Zai
90.3
1.7K
205K
¥7.2 / ¥23输入/输出
32
Aqwen3-max-preview
qwen3-max-previewAlibaba
89.9
1.3K
262K
¥6.2 / ¥24.8输入/输出
33
Aqwen3-235b-a22b-thinking-2507
qwen3-235b-a22b-thinking-2507Alibaba
89.6
411
131K
¥2.07 / ¥8.26输入/输出
34
Ggemini-2.5-pro
gemini-2.5-proGoogle
89.3
7.5K
1.05M
¥9 / ¥72输入/输出
35
Aqwen3.6-plus
qwen3.6-plusAlibaba
89.0
1.6K
1M
¥3.6 / ¥21.6输入/输出
36
D
deepseek-v4-pro
deepseek-v4-proDeepseek
88.6
1.6K
1M
¥3.13 / ¥6.26输入/输出
37
M
muse-spark
muse-sparkMeta
88.3
1.1K
-
-
38
M
longcat-flash-chat-2602-exp
longcat-flash-chat-2602-expMeituan
88.0
2.1K
128K
¥1.08 / ¥10.8输入/输出
39
Ggemma-4-31b
gemma-4-31bGoogle
87.7
437
262K
¥3.24 / ¥7.2输入/输出
40
X
grok-4.20-multi-agent-beta-0309
grok-4.20-multi-agent-beta-0309Xai
87.3
2.5K
2M
¥14.4 / ¥43.2输入/输出
41
Ggemma-4-26b-a4b
gemma-4-26b-a4bGoogle
87.0
406
262K
¥0.94 / ¥2.88输入/输出
42
Ddeepseek-v4-pro-thinking
deepseek-v4-pro-thinkingDeepseek
86.7
1.4K
1M
¥3.13 / ¥6.26输入/输出
43
Ogpt-5.2-high
gpt-5.2-highOpenai
86.4
3.5K
400K
¥12.6 / ¥101输入/输出
44
Aqwen3-vl-235b-a22b-instruct
qwen3-vl-235b-a22b-instructAlibaba
86.0
563
128K
¥2.16 / ¥8.64输入/输出
45
Aamazon-nova-experimental-chat-26-01-10
amazon-nova-experimental-chat-26-01-10Amazon
85.7
253
-
-
46
Aqwen3-235b-a22b-instruct-2507
qwen3-235b-a22b-instruct-2507Alibaba
85.4
5.7K
128K
¥2.09 / ¥8.23输入/输出
47
B
dola-seed-2.0-pro
dola-seed-2.0-proBytedance
85.1
3.2K
-
-
48
Ogpt-5.4-mini-high
gpt-5.4-mini-highOpenai
84.7
2.3K
400K
¥5.4 / ¥32.4输入/输出
49
Xgrok-4.20-beta-0309-reasoning
grok-4.20-beta-0309-reasoningXai
84.4
2.6K
2M
¥14.4 / ¥43.2输入/输出
50
Ogpt-5.1
gpt-5.1Openai
84.1
2.7K
400K
¥9 / ¥72输入/输出
51
Mkimi-k2.5-instant
kimi-k2.5-instantMoonshot
83.8
565
262K
¥4.32 / ¥21.6输入/输出
52
M
minimax-m2.1-preview
minimax-m2.1-previewMinimax
83.4
1.1K
205K
¥0 / ¥0输入/输出
53
Ogpt-5.2-chat-latest-20260210
gpt-5.2-chat-latest-20260210Openai
83.1
2.7K
400K
¥12.6 / ¥101输入/输出
54
Aclaude-haiku-4-5-20251001
claude-haiku-4-5-20251001Anthropic
82.8
5.4K
200K
¥7.2 / ¥36输入/输出
55
Aclaude-opus-4-1-20250805-thinking-16k
claude-opus-4-1-20250805-thinking-16kAnthropic
82.5
2.3K
200K
¥108 / ¥540输入/输出
56
Ddeepseek-v4-flash
deepseek-v4-flashDeepseek
82.1
1.5K
1M
¥1.01 / ¥2.02输入/输出
57
Mkimi-k2-thinking-turbo
kimi-k2-thinking-turboMoonshot
81.8
4.1K
262K
¥17.3 / ¥72输入/输出
58
Mlongcat-flash-chat
longcat-flash-chatMeituan
81.5
515
128K
¥1.08 / ¥10.8输入/输出
59
Aamazon-nova-experimental-chat-11-10
amazon-nova-experimental-chat-11-10Amazon
81.2
1.5K
-
-
60
Ddeepseek-v4-flash-thinking
deepseek-v4-flash-thinkingDeepseek
80.8
1.6K
1M
¥1.01 / ¥2.02输入/输出
61
Ddeepseek-v3.2
deepseek-v3.2Deepseek
80.5
3K
128K
¥2.09 / ¥3.1输入/输出
62
Zglm-4.6
glm-4.6Zai
80.2
1.9K
205K
¥4.32 / ¥15.8输入/输出
63
Mminimax-m2.7
minimax-m2.7Minimax
79.9
1.9K
205K
¥0 / ¥0输入/输出
64
Aqwen3.5-122b-a10b
qwen3.5-122b-a10bAlibaba
79.5
2.3K
262K
¥2.88 / ¥23输入/输出
65
Ogpt-5.2
gpt-5.2Openai
79.2
3.7K
400K
¥12.6 / ¥101输入/输出
66
Ddeepseek-v3.2-thinking
deepseek-v3.2-thinkingDeepseek
78.9
2.5K
128K
¥2.09 / ¥3.1输入/输出
67
TE
hunyuan-hy3-preview
hunyuan-hy3-previewTencent
78.6
575
256K
¥0 / ¥0输入/输出
68
Zglm-4.5
glm-4.5Zai
78.2
1.1K
131K
¥4.32 / ¥15.8输入/输出
69
Bernie-5.0-0110
ernie-5.0-0110Baidu
77.9
2.6K
128K
¥7.92 / ¥14.4输入/输出
70
Ggemini-2.5-flash
gemini-2.5-flashGoogle
77.6
7.6K
1.05M
¥2.16 / ¥18输入/输出
71
Aqwen3.5-27b
qwen3.5-27bAlibaba
77.3
2.2K
262K
¥2.16 / ¥17.3输入/输出
72
Xgrok-4.20-beta1
grok-4.20-beta1Xai
76.9
2K
2M
¥14.4 / ¥43.2输入/输出
73
Ggemini-3-flash (thinking-minimal)
gemini-3-flash (thinking-minimal)Google
76.6
4.2K
1.05M
¥3.6 / ¥21.6输入/输出
74
Aclaude-opus-4-1-20250805
claude-opus-4-1-20250805Anthropic
76.3
3.9K
200K
¥108 / ¥540输入/输出
75
MImimo-v2-flash (non-thinking)
mimo-v2-flash (non-thinking)Xiaomi
76.0
3.4K
262K
¥0.72 / ¥2.16输入/输出
76
Ggemini-2.5-flash-preview-09-2025
gemini-2.5-flash-preview-09-2025Google
75.6
1.6K
1M
¥2.16 / ¥18输入/输出
77
Bernie-5.0-preview-1022
ernie-5.0-preview-1022Baidu
75.3
276
128K
¥7.92 / ¥14.4输入/输出
78
Xgrok-4.1-thinking
grok-4.1-thinkingXai
75.0
4.4K
200K
¥14.4 / ¥72输入/输出
79
Aqwen3-vl-235b-a22b-thinking
qwen3-vl-235b-a22b-thinkingAlibaba
74.7
378
131K
¥2.06 / ¥8.26输入/输出
80
Bernie-5.0-preview-1203
ernie-5.0-preview-1203Baidu
74.4
672
128K
¥7.92 / ¥14.4输入/输出
81
MImimo-v2-omni
mimo-v2-omniXiaomi
74.0
283
262K
¥2.88 / ¥14.4输入/输出
82
Ogpt-5-high
gpt-5-highOpenai
73.7
1.6K
400K
¥9 / ¥72输入/输出
83
Zglm-4.7
glm-4.7Zai
73.4
717
205K
¥0 / ¥0输入/输出
84
ST
step-3.5-flash
step-3.5-flashStepfun
73.1
2.7K
256K
¥0.69 / ¥2.07输入/输出
85
Xgrok-3-preview-02-24
grok-3-preview-02-24Xai
72.7
1.5K
1M
¥9 / ¥18输入/输出
86
Xgrok-4-0709
grok-4-0709Xai
72.4
2K
256K
¥21.6 / ¥108输入/输出
87
Ddeepseek-v3.2-exp-thinking
deepseek-v3.2-exp-thinkingDeepseek
72.1
396
128K
¥0 / ¥0输入/输出
88
MA
mistral-large-3
mistral-large-3Mistral
71.8
3K
262K
¥3.6 / ¥10.8输入/输出
89
Xgrok-4.1
grok-4.1Xai
71.4
4.4K
200K
¥14.4 / ¥72输入/输出
90
Aamazon-nova-experimental-chat-12-10
amazon-nova-experimental-chat-12-10Amazon
71.1
242
-
-
91
Ogpt-5.5-instant
gpt-5.5-instantOpenai
70.8
2.4K
400K
¥9 / ¥72输入/输出
92
MAmistral-medium-2508
mistral-medium-2508Mistral
70.5
5.8K
262K
¥2.88 / ¥14.4输入/输出
93
Aqwen3.5-flash
qwen3.5-flashAlibaba
70.1
2.4K
1M
¥1.24 / ¥12.4输入/输出
94
Aqwen3-next-80b-a3b-instruct
qwen3-next-80b-a3b-instructAlibaba
69.8
1K
131K
¥1.04 / ¥4.13输入/输出
95
Aqwen3.5-35b-a3b
qwen3.5-35b-a3bAlibaba
69.5
2.4K
262K
¥1.8 / ¥14.4输入/输出
96
Xgrok-4-fast-reasoning
grok-4-fast-reasoningXai
69.2
862
2M
¥1.44 / ¥3.6输入/输出
97
Ogpt-5-chat
gpt-5-chatOpenai
68.8
1.5K
400K
¥9 / ¥72输入/输出
98
Xgrok-4-fast-chat
grok-4-fast-chatXai
68.5
298
2M
¥1.44 / ¥3.6输入/输出
99
MImimo-v2-flash (thinking)
mimo-v2-flash (thinking)Xiaomi
68.2
711
262K
¥0.72 / ¥2.16输入/输出
100
Ddeepseek-v3.1
deepseek-v3.1Deepseek
67.9
726
128K
¥1.44 / ¥5.04输入/输出
101
Xgrok-4-1-fast-reasoning
grok-4-1-fast-reasoningXai
67.5
3.8K
2M
¥1.44 / ¥3.6输入/输出
102
Ggemini-3.1-flash-lite-preview
gemini-3.1-flash-lite-previewGoogle
67.2
3K
1.05M
¥1.8 / ¥10.8输入/输出
103
Ddeepseek-v3.1-thinking
deepseek-v3.1-thinkingDeepseek
66.9
529
128K
¥1.44 / ¥5.04输入/输出
104
NV
nvidia-nemotron-3-super-120b-a12b
nvidia-nemotron-3-super-120b-a12bNvidia
66.6
594
262K
¥1.44 / ¥5.76输入/输出
105
Ddeepseek-v3.2-exp
deepseek-v3.2-expDeepseek
66.2
610
128K
¥0 / ¥0输入/输出
106
Oo3-2025-04-16
o3-2025-04-16Openai
65.9
3K
200K
¥14.4 / ¥57.6输入/输出
107
Ochatgpt-4o-latest-20250326
chatgpt-4o-latest-20250326Openai
65.6
4.3K
128K
¥18 / ¥72输入/输出
108
Aqwen3-max-2025-09-23
qwen3-max-2025-09-23Alibaba
65.3
449
258K
¥6.19 / ¥24.7输入/输出
109
Ddeepseek-r1-0528
deepseek-r1-0528Deepseek
64.9
881
164K
¥3.6 / ¥15.5输入/输出
110
Ogpt-5.3-chat-latest
gpt-5.3-chat-latestOpenai
64.6
2.6K
128K
¥12.6 / ¥101输入/输出
111
Ogpt-5.4-nano-high
gpt-5.4-nano-highOpenai
64.3
2.3K
400K
¥1.44 / ¥9输入/输出
112
Ogpt-4.5-preview-2025-02-27
gpt-4.5-preview-2025-02-27Openai
64.0
608
8.19K
¥216 / ¥432输入/输出
113
Aqwen3-30b-a3b-instruct-2507
qwen3-30b-a3b-instruct-2507Alibaba
63.6
1.1K
262K
¥2.16 / ¥3.6输入/输出
114
Xgrok-3-mini-high
grok-3-mini-highXai
63.3
921
128K
¥0 / ¥0输入/输出
115
TEhunyuan-t1-20250711
hunyuan-t1-20250711Tencent
63.0
215
131K
¥0 / ¥0输入/输出
116
Aamazon-nova-experimental-chat-10-20
amazon-nova-experimental-chat-10-20Amazon
62.7
652
-
-
117
Xgrok-4.3
grok-4.3Xai
62.3
1.5K
1M
¥9 / ¥18输入/输出
118
Aclaude-opus-4-20250514-thinking-16k
claude-opus-4-20250514-thinking-16kAnthropic
62.0
1.7K
200K
¥108 / ¥540输入/输出
119
Ogpt-5-mini-high
gpt-5-mini-highOpenai
61.7
1.1K
400K
¥1.8 / ¥14.4输入/输出
120
Mminimax-m2.5
minimax-m2.5Minimax
61.4
3K
205K
¥0 / ¥0输入/输出
121
Ggemini-2.5-flash-lite-preview-09-2025-no-thinking
gemini-2.5-flash-lite-preview-09-2025-no-thinkingGoogle
61.0
2.5K
1.05M
¥0.72 / ¥2.88输入/输出
122
Aclaude-sonnet-4-20250514-thinking-32k
claude-sonnet-4-20250514-thinking-32kAnthropic
60.7
1.7K
200K
¥21.6 / ¥108输入/输出
123
Aclaude-opus-4-20250514
claude-opus-4-20250514Anthropic
60.4
2.2K
200K
¥108 / ¥540输入/输出
124
Zglm-4.5-air
glm-4.5-airZai
60.1
1.4K
131K
¥0 / ¥0输入/输出
125
Aqwen3-235b-a22b-no-thinking
qwen3-235b-a22b-no-thinkingAlibaba
59.7
1.9K
131K
¥2.07 / ¥8.26输入/输出
126
Aqwen3-next-80b-a3b-thinking
qwen3-next-80b-a3b-thinkingAlibaba
59.4
620
131K
¥1.04 / ¥10.3输入/输出
127
NVnvidia-nemotron-3-nano-30b-a3b-bf16
nvidia-nemotron-3-nano-30b-a3b-bf16Nvidia
59.1
944
131K
¥0 / ¥0输入/输出
128
Anova-2-lite
nova-2-liteAmazon
58.8
713
128K
¥2.38 / ¥19.8输入/输出
129
Mkimi-k2-0905-preview
kimi-k2-0905-previewMoonshot
58.4
555
262K
¥4.32 / ¥18输入/输出
130
Ggemini-2.5-flash-lite-preview-06-17-thinking
gemini-2.5-flash-lite-preview-06-17-thinkingGoogle
58.1
1.6K
65.5K
¥0.72 / ¥2.88输入/输出
131
Ogpt-4.1-2025-04-14
gpt-4.1-2025-04-14Openai
57.8
2.5K
1.05M
¥14.4 / ¥57.6输入/输出
132
Oo3-mini-high
o3-mini-highOpenai
57.5
847
200K
¥7.92 / ¥31.7输入/输出
133
Aqwen3-32b
qwen3-32bAlibaba
57.1
236
131K
¥2.07 / ¥8.26输入/输出
134
Zglm-4.7-flash
glm-4.7-flashZai
56.8
795
200K
¥0 / ¥0输入/输出
135
Oo1-2024-12-17
o1-2024-12-17Openai
56.5
1.3K
128K
¥108 / ¥432输入/输出
136
UNtrinity-large-thinking
-
56.2
2.1K
262K
¥1.8 / ¥6.48输入/输出
137
Xgrok-3-mini-beta
grok-3-mini-betaXai
55.8
1.2K
1M
¥9 / ¥18输入/输出
138
Ogpt-oss-120b
gpt-oss-120bOpenai
55.5
1.3K
131K
¥1.08 / ¥4.32输入/输出
139
Zglm-4.5v
glm-4.5vZai
55.2
196
64K
¥4.32 / ¥13输入/输出
140
IA
mercury-2
mercury-2Inception Ai
54.9
229
128K
¥1.8 / ¥5.4输入/输出
141
AG
ling-flash-2.0
ling-flash-2.0Ant Group
54.5
341
131K
¥1.01 / ¥4.1输入/输出
142
Ddeepseek-v3-0324
deepseek-v3-0324Deepseek
54.2
2.3K
75K
¥1.44 / ¥5.76输入/输出
143
UNtrinity-large-preview
-
53.9
2.4K
262K
¥1.8 / ¥6.48输入/输出
144
Aqwen3-235b-a22b
qwen3-235b-a22bAlibaba
53.6
1.3K
131K
¥2.07 / ¥8.26输入/输出
145
Mkimi-k2-0711-preview
kimi-k2-0711-previewMoonshot
53.2
1.5K
131K
¥4.32 / ¥18输入/输出
146
Oo4-mini-2025-04-16
o4-mini-2025-04-16Openai
52.9
2.3K
200K
¥7.92 / ¥31.7输入/输出
147
MAmistral-medium-2505
mistral-medium-2505Mistral
52.6
1.8K
262K
¥2.88 / ¥14.4输入/输出
148
Ggemini-2.0-flash-001
gemini-2.0-flash-001Google
52.3
2.2K
1.05M
¥1.08 / ¥4.32输入/输出
149
AGring-flash-2.0
ring-flash-2.0Ant Group
51.9
331
131K
¥1.01 / ¥4.1输入/输出
150
TEhunyuan-turbos-20250416
hunyuan-turbos-20250416Tencent
51.6
581
131K
¥0 / ¥0输入/输出
151
Ogpt-4.1-mini-2025-04-14
gpt-4.1-mini-2025-04-14Openai
51.3
2K
1.05M
¥2.88 / ¥11.5输入/输出
152
Oo1-preview
o1-previewOpenai
51.0
2K
128K
¥108 / ¥432输入/输出
153
Ddeepseek-r1
deepseek-r1Deepseek
50.6
848
164K
¥5.04 / ¥18输入/输出
154
Aqwen3-coder-480b-a35b-instruct
qwen3-coder-480b-a35b-instructAlibaba
50.3
1.3K
262K
¥6.2 / ¥24.8输入/输出
155
Aqwen2.5-max
qwen2.5-maxAlibaba
50.0
1.7K
32K
¥11.5 / ¥46输入/输出
156
Aclaude-sonnet-4-20250514
claude-sonnet-4-20250514Anthropic
49.7
2K
200K
¥21.6 / ¥108输入/输出
157
Oo3-mini
o3-miniOpenai
49.4
2.9K
200K
¥7.92 / ¥31.7输入/输出
158
STstep-3
step-3Stepfun
49.0
259
65.5K
¥1.8 / ¥4.68输入/输出
159
Aqwen-plus-0125
qwen-plus-0125Alibaba
48.7
358
1M
¥0.83 / ¥2.07输入/输出
160
Aqwq-32b
qwq-32bAlibaba
48.4
1.2K
131K
¥2.07 / ¥6.2输入/输出
161
Mminimax-m2
minimax-m2Minimax
48.1
258
197K
¥0 / ¥0输入/输出
162
Ogpt-5-nano-high
gpt-5-nano-highOpenai
47.7
324
400K
¥0.36 / ¥2.88输入/输出
163
NVnvidia-llama-3.3-nemotron-super-49b-v1.5
nvidia-llama-3.3-nemotron-super-49b-v1.5Nvidia
47.4
180
131K
¥2.88 / ¥2.88输入/输出
164
Aclaude-3-7-sonnet-20250219-thinking-32k
claude-3-7-sonnet-20250219-thinking-32kAnthropic
47.1
1.9K
-
-
165
UNintellect-3
-
46.8
309
131K
¥1.44 / ¥7.92输入/输出
166
Mminimax-m1
minimax-m1Minimax
46.4
1.6K
1M
¥0.95 / ¥9.03输入/输出
167
Oo1-mini
o1-miniOpenai
46.1
3.2K
128K
¥7.92 / ¥31.7输入/输出
168
Aqwen3-30b-a3b
qwen3-30b-a3bAlibaba
45.8
1.3K
128K
¥0.79 / ¥7.78输入/输出
169
AI
olmo-3.1-32b-instruct
olmo-3.1-32b-instructAllenai
45.5
752
200K
¥14.4 / ¥57.6输入/输出
170
IB
granite-4.1-8b
granite-4.1-8bIbm
45.1
353
131K
¥0.36 / ¥0.72输入/输出
171
Ddeepseek-v3
deepseek-v3Deepseek
44.8
1.2K
128K
¥0 / ¥0输入/输出
172
Ggemini-2.0-flash-lite-preview-02-05
gemini-2.0-flash-lite-preview-02-05Google
44.5
1.2K
1.05M
¥0.54 / ¥2.16输入/输出
173
Ggemma-3-27b-it
gemma-3-27b-itGoogle
44.2
2.2K
128K
¥2.15 / ¥2.15输入/输出
174
Aclaude-3-7-sonnet-20250219
claude-3-7-sonnet-20250219Anthropic
43.8
2.1K
200K
¥21.6 / ¥108输入/输出
175
STstep-1o-turbo-202506
step-1o-turbo-202506Stepfun
43.5
472
-
-
176
MAmistral-small-2506
mistral-small-2506Mistral
43.2
840
262K
¥2.88 / ¥14.4输入/输出
177
CO
command-a-03-2025
command-a-03-2025Cohere
42.9
2.8K
256K
¥18 / ¥72输入/输出
178
Aqwen2.5-plus-1127
qwen2.5-plus-1127Alibaba
42.5
664
-
-
179
AIolmo-3-32b-think
olmo-3-32b-thinkAllenai
42.2
275
128K
¥2.16 / ¥3.24输入/输出
180
UNyi-lightning
-
41.9
1.5K
12K
¥1.44 / ¥1.44输入/输出
181
AIolmo-3.1-32b-think
olmo-3.1-32b-thinkAllenai
41.6
503
200K
¥14.4 / ¥57.6输入/输出
182
Zglm-4-plus-0111
glm-4-plus-0111Zai
41.2
354
128K
¥72 / ¥72输入/输出
183
STstep-2-16k-exp-202412
step-2-16k-exp-202412Stepfun
40.9
310
16.4K
¥37.5 / ¥118输入/输出
184
Ggemini-1.5-pro-002
gemini-1.5-pro-002Google
40.6
3.3K
-
-
185
TEhunyuan-large-2025-02-10
hunyuan-large-2025-02-10Tencent
40.3
228
-
-
186
Ogpt-4.1-nano-2025-04-14
gpt-4.1-nano-2025-04-14Openai
39.9
328
1.05M
¥14.4 / ¥57.6输入/输出
187
UNathene-v2-chat
-
39.6
1.5K
-
-
188
Ddeepseek-v2.5-1210
deepseek-v2.5-1210Deepseek
39.3
441
1M
¥1.01 / ¥2.02输入/输出
189
Aclaude-3-5-sonnet-20241022
claude-3-5-sonnet-20241022Anthropic
39.0
5K
200K
¥21.6 / ¥108输入/输出
190
Ogpt-oss-20b
gpt-oss-20bOpenai
38.6
489
131K
¥0.32 / ¥1.3输入/输出
191
Mllama-4-maverick-17b-128e-instruct
llama-4-maverick-17b-128e-instructMeta
38.3
2K
1M
¥1.8 / ¥6.26输入/输出
192
MAmistral-small-3.1-24b-instruct-2503
mistral-small-3.1-24b-instruct-2503Mistral
38.0
1.6K
262K
¥2.88 / ¥14.4输入/输出
193
Xgrok-2-2024-08-13
grok-2-2024-08-13Xai
37.7
3.5K
1M
¥9 / ¥18输入/输出
194
TEhunyuan-large-vision
hunyuan-large-visionTencent
37.3
287
-
-
195
Zglm-4-plus
glm-4-plusZai
37.0
1.6K
128K
¥54 / ¥54输入/输出
196
Ogpt-4o-2024-05-13
gpt-4o-2024-05-13Openai
36.7
5.9K
128K
¥36 / ¥108输入/输出
197
Ggemma-3-12b-it
gemma-3-12b-itGoogle
36.4
186
128K
¥1.96 / ¥1.96输入/输出
198
Aqwen-max-0919
qwen-max-0919Alibaba
36.0
1K
131K
¥2.48 / ¥9.91输入/输出
199
TEhunyuan-standard-2025-02-10
hunyuan-standard-2025-02-10Tencent
35.7
207
-
-
200
Ggemma-3n-e4b-it
gemma-3n-e4b-itGoogle
35.4
1.1K
128K
¥0 / ¥0输入/输出
201
Aclaude-3-5-sonnet-20240620
claude-3-5-sonnet-20240620Anthropic
35.1
4.3K
200K
¥21.6 / ¥108输入/输出
202
Aqwen2.5-72b-instruct
qwen2.5-72b-instructAlibaba
34.7
2.4K
131K
¥4.13 / ¥12.4输入/输出
203
Ggemini-1.5-pro-001
gemini-1.5-pro-001Google
34.4
3.9K
-
-
204
Mllama-3.1-405b-instruct-fp8
llama-3.1-405b-instruct-fp8Meta
34.1
3.1K
128K
¥0 / ¥0输入/输出
205
NVllama-3.1-nemotron-70b-instruct
llama-3.1-nemotron-70b-instructNvidia
33.8
453
128K
¥0 / ¥0输入/输出
206
IBibm-granite-h-small
ibm-granite-h-smallIbm
33.4
283
-
-
207
Ogpt-4o-2024-08-06
gpt-4o-2024-08-06Openai
33.1
2.3K
128K
¥18 / ¥72输入/输出
208
Ddeepseek-v2.5
deepseek-v2.5Deepseek
32.8
1.5K
1M
¥1.01 / ¥2.02输入/输出
209
Mllama-4-scout-17b-16e-instruct
llama-4-scout-17b-16e-instructMeta
32.5
1.5K
128K
¥1.44 / ¥5.62输入/输出
210
Xgrok-2-mini-2024-08-13
grok-2-mini-2024-08-13Xai
32.1
2.8K
1M
¥9 / ¥18输入/输出
211
Ogpt-4o-mini-2024-07-18
gpt-4o-mini-2024-07-18Openai
31.8
3.5K
128K
¥1.08 / ¥4.32输入/输出
212
Ggemini-1.5-flash-002
gemini-1.5-flash-002Google
31.5
2.1K
2M
¥0.54 / ¥2.2输入/输出
213
MAmistral-large-2407
mistral-large-2407Mistral
31.2
2.5K
131K
¥14.4 / ¥43.2输入/输出
214
Mllama-3.1-405b-instruct-bf16
llama-3.1-405b-instruct-bf16Meta
30.8
2.1K
128K
¥0 / ¥0输入/输出
215
UNathene-70b-0725
-
30.5
867
-
-
216
Mllama-3.3-70b-instruct
llama-3.3-70b-instructMeta
30.2
2.9K
128K
¥0 / ¥0输入/输出
217
Ggemma-3-4b-it
gemma-3-4b-itGoogle
29.9
208
128K
¥1.44 / ¥1.44输入/输出
218
Ogpt-4-turbo-2024-04-09
gpt-4-turbo-2024-04-09Openai
29.5
5.2K
128K
¥72 / ¥216输入/输出
219
Aclaude-3-opus-20240229
claude-3-opus-20240229Anthropic
29.2
10.4K
200K
¥108 / ¥540输入/输出
220
Aqwen2.5-coder-32b-instruct
qwen2.5-coder-32b-instructAlibaba
28.9
267
131K
¥2.07 / ¥6.2输入/输出
221
MAmagistral-medium-2506
magistral-medium-2506Mistral
28.6
553
128K
¥14.4 / ¥36输入/输出
222
UNjamba-1.5-large
-
28.2
331
256K
¥0 / ¥0输入/输出
223
Ggemini-advanced-0514
gemini-advanced-0514Google
27.9
2.4K
-
-
224
UNreka-core-20240904
-
27.6
458
-
-
225
Ogpt-4-1106-preview
gpt-4-1106-previewOpenai
27.3
4.2K
8.19K
¥216 / ¥432输入/输出
226
Aamazon-nova-pro-v1.0
amazon-nova-pro-v1.0Amazon
26.9
1.4K
300K
¥5.76 / ¥23输入/输出
227
Mllama-3.1-70b-instruct
llama-3.1-70b-instructMeta
26.6
2.9K
131K
¥2.88 / ¥2.88输入/输出
228
MAmistral-large-2411
mistral-large-2411Mistral
26.3
1.5K
128K
¥14.4 / ¥43.2输入/输出
229
Aclaude-3-5-haiku-20241022
claude-3-5-haiku-20241022Anthropic
26.0
3.5K
200K
¥5.76 / ¥28.8输入/输出
230
Ogpt-4-0125-preview
gpt-4-0125-previewOpenai
25.6
4.5K
8.19K
¥216 / ¥432输入/输出
231
M
phi-4
phi-4Microsoft
25.3
1.1K
128K
¥0.9 / ¥3.6输入/输出
232
MAmistral-small-24b-instruct-2501
mistral-small-24b-instruct-2501Mistral
25.0
754
262K
¥2.88 / ¥14.4输入/输出
233
Aamazon-nova-lite-v1.0
amazon-nova-lite-v1.0Amazon
24.7
1.1K
300K
¥0.43 / ¥1.73输入/输出
234
Ggemini-1.5-flash-001
gemini-1.5-flash-001Google
24.4
3.2K
2M
¥0.54 / ¥2.2输入/输出
235
Ggemini-1.5-flash-8b-001
gemini-1.5-flash-8b-001Google
24.0
2.1K
2M
¥0.54 / ¥2.2输入/输出
236
Aamazon-nova-micro-v1.0
amazon-nova-micro-v1.0Amazon
23.7
1.1K
128K
¥0.25 / ¥1.01输入/输出
237
COc4ai-aya-expanse-32b
c4ai-aya-expanse-32bCohere
23.4
1.8K
-
-
238
UNreka-flash-20240904
-
23.1
493
65.5K
¥0.72 / ¥1.44输入/输出
239
Ddeepseek-coder-v2
deepseek-coder-v2Deepseek
22.7
769
1M
¥1.01 / ¥2.02输入/输出
240
Zglm-4-0520
glm-4-0520Zai
22.4
514
128K
¥108 / ¥108输入/输出
241
COcommand-r-plus-08-2024
command-r-plus-08-2024Cohere
22.1
524
128K
¥18 / ¥72输入/输出
242
Aclaude-3-sonnet-20240229
claude-3-sonnet-20240229Anthropic
21.8
5.6K
200K
¥21.6 / ¥108输入/输出
243
Ggemma-2-27b-it
gemma-2-27b-itGoogle
21.4
4K
8.19K
¥0.58 / ¥0.58输入/输出
244
NVnemotron-4-340b-instruct
nemotron-4-340b-instructNvidia
21.1
1K
-
-
245
Aqwen2-72b-instruct
qwen2-72b-instructAlibaba
20.8
1.8K
131K
¥4.13 / ¥12.4输入/输出
246
MAministral-8b-2410
ministral-8b-2410Mistral
20.5
332
128K
¥0.72 / ¥0.72输入/输出
247
NVllama-3.1-nemotron-51b-instruct
llama-3.1-nemotron-51b-instructNvidia
20.1
265
128K
¥0 / ¥0输入/输出
248
UNgemma-2-9b-it-simpo
-
19.8
369
8.19K
¥1.44 / ¥1.44输入/输出
249
COcommand-r-plus
command-r-plusCohere
19.5
4K
128K
¥18 / ¥72输入/输出
250
COc4ai-aya-expanse-8b
c4ai-aya-expanse-8bCohere
19.2
600
-
-
251
UNinternlm2_5-20b-chat
-
18.8
604
-
-
252
Mllama-3-70b-instruct
llama-3-70b-instructMeta
18.5
8K
8.19K
¥3.67 / ¥5.33输入/输出
253
Ogpt-4-0314
gpt-4-0314Openai
18.2
2.2K
8.19K
¥216 / ¥432输入/输出
254
Aclaude-3-haiku-20240307
claude-3-haiku-20240307Anthropic
17.9
6.3K
200K
¥1.8 / ¥9输入/输出
255
Ggemma-2-9b-it
gemma-2-9b-itGoogle
17.5
2.8K
8.19K
¥1.44 / ¥1.44输入/输出
256
Aqwen1.5-110b-chat
qwen1.5-110b-chatAlibaba
17.2
1.4K
-
-
257
UNyi-1.5-34b-chat
-
16.9
1K
-
-
258
Mllama-3.1-8b-instruct
llama-3.1-8b-instructMeta
16.6
2.6K
131K
¥0.79 / ¥0.79输入/输出
259
IBgranite-3.1-8b-instruct
granite-3.1-8b-instructIbm
16.2
237
-
-
260
COcommand-r-08-2024
command-r-08-2024Cohere
15.9
604
128K
¥18 / ¥72输入/输出
261
Aqwen1.5-72b-chat
qwen1.5-72b-chatAlibaba
15.6
1.8K
-
-
262
UNjamba-1.5-mini
-
15.3
332
256K
¥0 / ¥0输入/输出
263
Aqwq-32b-preview
qwq-32b-previewAlibaba
14.9
232
131K
¥2.07 / ¥6.2输入/输出
264
IBgranite-3.1-2b-instruct
granite-3.1-2b-instructIbm
14.6
225
-
-
265
Ogpt-4-0613
gpt-4-0613Openai
14.3
3.6K
8.19K
¥216 / ¥432输入/输出
266
Aqwen1.5-32b-chat
qwen1.5-32b-chatAlibaba
14.0
1.2K
-
-
267
MAmistral-medium
mistral-mediumMistral
13.6
1.4K
262K
¥2.88 / ¥14.4输入/输出
268
MAmistral-large-2402
mistral-large-2402Mistral
13.3
2.9K
262K
¥2.88 / ¥14.4输入/输出
269
UNreka-flash-21b-20240226-online
-
13.0
830
-
-
270
Mllama-3-8b-instruct
llama-3-8b-instructMeta
12.7
5.4K
8.19K
¥0.29 / ¥0.29输入/输出
271
MAmixtral-8x22b-instruct-v0.1
mixtral-8x22b-instruct-v0.1Mistral
12.3
2.6K
64K
¥14.4 / ¥43.2输入/输出
272
Mphi-3-medium-4k-instruct
phi-3-medium-4k-instructMicrosoft
12.0
1.1K
4.1K
¥1.22 / ¥4.9输入/输出
273
COcommand-r
command-rCohere
11.7
2.8K
128K
¥18 / ¥72输入/输出
274
UNreka-flash-21b-20240226
-
11.4
1.3K
-
-
275
Ggemma-2-2b-it
gemma-2-2b-itGoogle
11.0
2.5K
128K
¥0 / ¥0输入/输出
276
Aqwen1.5-14b-chat
qwen1.5-14b-chatAlibaba
10.7
944
-
-
277
Mllama-3.2-3b-instruct
llama-3.2-3b-instructMeta
10.4
499
131K
¥0.22 / ¥0.35输入/输出
278
MAmixtral-8x7b-instruct-v0.1
mixtral-8x7b-instruct-v0.1Mistral
10.1
3.2K
32K
¥5.04 / ¥5.04输入/输出
279
IBgranite-3.0-8b-instruct
granite-3.0-8b-instructIbm
9.7
345
-
-
280
UNstarling-lm-7b-beta
-
9.4
952
200K
¥5.4 / ¥18.7输入/输出
281
UNdbrx-instruct-preview
-
9.1
1.7K
-
-
282
Ogpt-3.5-turbo-1106
gpt-3.5-turbo-1106Openai
8.8
437
16.4K
¥7.2 / ¥14.4输入/输出
283
Mphi-3-small-8k-instruct
phi-3-small-8k-instructMicrosoft
8.4
895
8.19K
¥1.08 / ¥4.32输入/输出
284
Ogpt-3.5-turbo-0125
gpt-3.5-turbo-0125Openai
8.1
3.2K
16.4K
¥3.6 / ¥10.8输入/输出
285
IBgranite-3.0-2b-instruct
granite-3.0-2b-instructIbm
7.8
407
-
-
286
UNzephyr-orpo-141b-A35b-v0.1
-
7.5
219
200K
¥108 / ¥432输入/输出
287
UNyi-34b-chat
-
7.1
559
-
-
288
Ggemini-pro-dev-api
gemini-pro-dev-apiGoogle
6.8
694
1.05M
¥14.4 / ¥86.4输入/输出
289
Aqwen1.5-7b-chat
qwen1.5-7b-chatAlibaba
6.5
231
-
-
290
UNopenchat-3.5-0106
-
6.2
611
-
-
291
Mphi-3-mini-4k-instruct-june-2024
phi-3-mini-4k-instruct-june-2024Microsoft
5.8
486
4.1K
¥0.94 / ¥3.74输入/输出
292
Mphi-3-mini-4k-instruct
phi-3-mini-4k-instructMicrosoft
5.5
936
4.1K
¥0.94 / ¥3.74输入/输出
293
UNopenchat-3.5
-
5.2
194
-
-
294
Ggemma-1.1-7b-it
gemma-1.1-7b-itGoogle
4.9
1.2K
-
-
295
Mllama-2-70b-chat
llama-2-70b-chatMeta
4.5
1.4K
-
-
296
Mllama-2-7b-chat
llama-2-7b-chatMeta
4.2
411
128K
¥4.03 / ¥48输入/输出
297
MAmistral-7b-instruct-v0.2
mistral-7b-instruct-v0.2Mistral
3.9
798
262K
¥2.88 / ¥14.4输入/输出
298
UNstarling-lm-7b-alpha
-
3.6
318
200K
¥5.4 / ¥18.7输入/输出
299
Mllama-2-13b-chat
llama-2-13b-chatMeta
3.2
539
-
-
300
UNvicuna-33b
-
2.9
484
-
-
301
UNsnowflake-arctic-instruct
-
2.6
1.7K
-
-
302
Mllama-3.2-1b-instruct
llama-3.2-1b-instructMeta
2.3
487
16.4K
¥0.07 / ¥0.08输入/输出
303
UNzephyr-7b-beta
-
1.9
201
-
-
304
Ggemma-7b-it
gemma-7b-itGoogle
1.6
340
-
-
305
UNvicuna-13b
-
1.3
321
-
-
306
Mphi-3-mini-128k-instruct
phi-3-mini-128k-instructMicrosoft
1.0
1.1K
128K
¥0.94 / ¥3.74输入/输出
307
Aqwen1.5-4b-chat
qwen1.5-4b-chatAlibaba
0.6
356
-
-
308
Ggemma-1.1-2b-it
gemma-1.1-2b-itGoogle
0.3
578
-
-
309
MAmistral-7b-instruct
mistral-7b-instructMistral
0.0
183
262K
¥2.88 / ¥14.4输入/输出
claude-opus-4-6 为什么排第一
claude-opus-4-6 当前排名第 1，厂商为 anthropic，百分制分数 100.0，样本量 3.1K。如果你的任务与本榜单口径一致，可以优先把它作为候选模型，再结合价格、上下文长度和可用性做最终选择。
不要只看第一名
先选择最接近真实任务的排行榜，再比较 Top 模型的分数和样本量。分数越高说明同一榜单内偏好表现越强；样本量越高通常结果更稳定。最后再结合价格、上下文长度、开源闭源和厂商可用性筛选。
继续比较相近能力
专家榜常见问题
专家榜排行榜看什么指标？
主要看排名、百分制分数、样本量和来源。分数用于快速比较同一榜单内模型表现，样本量用于判断结果稳定性。
为什么不同榜单不能直接混合成总分？
不同榜单的任务、样本和评测口径不同，模力榜默认只在同一榜单内排序，避免把写作、代码、图像等能力强行合并。
专家榜模型应该怎么选？
优先看与你任务最接近的榜单，再结合价格、上下文长度、开源闭源和厂商可用性。排名靠前不代表适合所有预算和部署方式。
榜单多久更新？
页面展示的是最新成功采集的公开榜单数据。当前优先使用 LMArena leaderboard dataset，并在页面来源中保留原始链接。