名次模型 综合智能 写代码 当助手
1 Claude Fable 5.2
64.6
—
—
2 Gemini 4 Carbon
60
—
—
3 Claude Opus 5.5
59.1
—
—
4 GPT-6 Astra
56.9
76.9
51
5 GPT-6.1 Sol
54.6
—
—
6 Claude Sonnet 5.5
54.5
—
—
7 Gemini 4 Hydrogen
53.9
—
—
8 Qwen3.8 Max
53.4
68.9
49.9
9 GPT-6.1 Terra
51.3
—
—
10 Gemini 4 Argon
50.4
—
—
11 Claude Fable 5.1
50
80.7
57.2
12 Kimi K3.1 Pro
46.5
—
—
13 GPT-6 Sol
44.2
—
—
14 Claude Opus 5
44.2
78
56.5
15 Muse Wonder
43.6
—
—
16 Muse Spark 1.3
36.6
75.8
55.5
17 Claude Haiku 5.5
36.5
—
—
18 Grok 4.7
35.5
—
—
19 Tencent Hy4
34.6
—
—
20 Qwen3.8 Max
33.7
76.2
56
21 GPT-5.3 Codex
32.5
—
—
22 GPT-5.6 Terra
32.3
76.7
43.2
23 Claude Opus 4.6
31.9
—
—
24 GLM-5.3
31.7
74.8
53.1
25 Grok 4.6
31
75.9
52.5
26 Gemini 3.8 Flash
30.8
76.3
40.2
27 Inkling 2 Small
30.6
—
—
28 GPT-5.2
30.4
—
—
29 MiMo-V2.6-Pro
29.9
—
—
30 Ling 3.1 Flash 1019
29.6
—
—
31 MiniMax-M3.1-Flash
29.1
—
—
32 Claude Opus 4.5
29.1
—
—
33 MiMo-V2-Pro
28.6
—
—
34 DeepSeek V4.1 Flash 1017
28.6
—
—
35 GPT-5.2 Codex
28.5
—
—
36 Step 5 Preview
28.5
—
—
37 Qwen3.6 Max Preview
28.4
—
—
38 Solar Pro 4
28.2
52.7
—
39 GLM-5
27.9
—
—
40 GPT-6 Luna
27.9
—
—
41 GLM-5.3-Flash
27.8
71.5
50.9
42 MiMo-V2.6-Flash
26.7
—
—
43 GLM-5-Turbo
26.6
—
—
44 Gemini 3 Flash Preview
26.3
—
—
45 Mistral Large 4
26.1
—
—
46 Grok 4.20 0309 v2
25.7
—
—
47 MiMo-V2-Omni-0327
25.1
—
—
48 GPT-5 Codex
24.9
—
—
49 Muse Spark Open
24.9
—
—
50 DeepSeek V4.1 Flash
24.7
—
—
51 Solar Mini 4
24.1
—
—
52 GPT-5.6 Luna
23.8
71.4
42.1
53 GPT-5.1 Codex
23.7
—
—
54 GLM 5V Turbo
23.5
—
—
55 DeepSeek V4 Pro 0813
23.1
68.8
41.3
56 Qwen3.5 27B
22.9
—
—
57 Claude 4.1 Opus
22.8
—
—
58 MiniMax-M2.5
22.8
—
—
59 Grok 4
22.5
—
—
60 Kimi K2 Thinking
22
—
—
61 o3-pro
21.9
—
—
62 Qwen3 Max Thinking
21.3
—
—
63 DeepSeek V4 Pro 0813
20.8
—
—
64 Qwen3.8 27B
20.1
68.1
45.8
65 MiniMax-M3
18.9
58.6
29.5
66 Nova 2.0 Lite
18.4
23
3.1
67 Beam 1.0
17.7
—
—
68 Kimi K3
17.4
72
29.3
69 Qwen3 Next 80B A3B
16.9
17.4
2.1
70 GPT-5 mini
16.8
15.6
6.8
71 Gemini 2.5 Pro
16.1
33.3
1.6
72 GPT-4.1 mini
14.8
20.2
1.8
73 Nemotron 3.5 Super VL
13.6
—
—
74 Command A+
13.1
27.8
1
75 Inkling Small
12.9
52.9
23.5
76 Nemotron 3 Super 120B A12B
12.8
37.7
1.7
77 Qwen3 235B A22B 2507
12.7
22.1
1.3
78 Inkling
12.5
52.1
22.5
79 Claude 3.5 Haiku
12.4
15.9
1.9
80 Mercury 2.5
12.3
—
—
81 Nemotron 3 Ultra 550B A55B
11.8
49.3
20.1
82 Gemini 3.5 Flash-Lite
11.7
49.3
14.3
83 Mercury 2
11.5
31.1
4
84 DeepSeek R1
11.4
24.6
1.1
85 Qwen3.5 9B
11.2
28.7
1.2
86 Ling 3.0 Tiny
11.1
26.5
4.4
87 o3-mini
11
16.3
0.9
88 Trinity Large Thinking
10.8
25.8
1
89 Qwen3 30B A3B 2507
9.8
12.1
0.9
90 DeepSeek V3 0324
9.7
21.2
0.8
91 GPT-4.1 nano
9.6
11.1
1.2
92 Llama 3.3 Instruct 70B
9.4
11.9
0.3
93 Llama 4 Maverick
9.3
16.3
0.6
94 Mistral Large 3
9.3
20.1
1
95 Qwen3 Coder Next
9.2
36.2
0.9
96 Mistral Medium 3.1
9.2
20.5
1.9
97 gpt-oss-20b
9
20.7
1.2
98 NVIDIA Nemotron 3 Nano 30B A3B
8.9
14.4
1
99 Devstral 2
8.6
31.3
2.3
100 DeepSeek V3
8.5
23
0.8
101 Solar Pro 3
7.8
16.2
1.4
102 Llama 3.1 Instruct 8B
7.6
5.4
0.5
103 Devstral Small 2
7.5
29.3
2
104 Claude 4.5 Haiku
7.4
43.9
8
105 Qwen3 32B
7.2
15.3
0.9
106 Mistral Small 3.1
7.1
26.3
1.2
107 Llama 4 Scout
6.5
8.2
0.5
108 Qwen3 14B
6.4
13.8
0.9
109 Muse Glimmer
6.3
49
8.5
110 Command A+
6.2
27.8
1
111 Gemini 3.1 Flash-Lite
6.2
34.7
1.6
112 Ministral 3 14B
6
14.4
1.1
113 Phi-4 Mini Instruct
6
3.8
0.3
114 Nemotron 3.5 Lightning
5.5
26.8
3.5
115 Ministral 3 8B
5.5
9.7
0.6
116 Qwen3 8B
5.2
9
0.8
117 Granite 4.2 8B
5.1
22.4
1.3
118 Gemma 3 27B Instruct
4.9
10.1
0.1
119 Ministral 3 3B
4.8
4.8
0.8
120 Mistral Medium 3.5
4.6
46.9
7
121 Gemma 4 31B
4.4
43.4
4.2
122 Mistral Small 4
4.3
26.6
0.8
123 Gemma 3 12B Instruct
3.8
5.8
0.1
124 gpt-oss-120b
3.8
30.4
3.7
125 North Mini Code
3.4
36.5
1.1
