跳到主内容
知行录
leaderboard.cn
Toggle Sidebar
搜索
⌘K
分析
前沿模型
历史趋势
本地模型
资料库
模型库
厂商
基准评测
数据源
显示模式
Toggle Sidebar
知行录
leaderboard.cn
正在加载页面内容
正在加载页面内容
首页
代码
覆盖软件工程、竞赛编程和终端任务。
更新于 07/16 02:55
模型
150 个 · 16 项基准
评分
SWE-bench Pro
Arena WebDev Rating
DeepSWE 1.1
CyberGym
Terminal Bench 2.1 (Terminus-2)
Artificial Analysis Intelligence Index v4.1
Artificial Analysis Coding Agent Index v1.1
SWE-Marathon
DeepSWE 1.0
SWE-bench Multilingual
SWE-bench Verified
NL2Repo
Claw Eval (pass^3)
CL-bench
PostTrainBench
ProgramBench
1
Claude Fable 5
100.0
80.4%
1652.7 Elo
70%
83.8%
88%
59.9 pts
77.2 pts
24%
66.1%
-
-
-
-
-
-
-
2
GPT-5.6 Sol
99.0
64.6%
1631.0 Elo
72.7%
84.5%
-
58.9 pts
80 pts
-
-
-
-
-
-
-
-
-
3
Claude Opus 4.8
98.6
69.2%
1560.7 Elo
59%
78.1%
85%
55.7 pts
72.5 pts
26%
55.8%
84.8%
88.6%
69.7%
72.1%
24.8%
37.2%
71.9%
4
GPT-5.6 Terra
95.3
63.4%
-
69.6%
81.8%
-
55 pts
77.4 pts
-
-
-
-
-
-
-
-
-
5
Claude Opus 4.7
94.4
64.3%
1558.9 Elo
-
73.1%
66.1%
-
-
16%
40.1%
80.5%
87.6%
58.2%
71.6%
22.9%
42.4%
52.1%
6
Grok 4.5
93.7
64.7%
1557.0 Elo
53%
-
-
-
-
29%
62%
-
-
-
-
-
-
-
7
GPT-5.5
93.6
59.4%
1501.8 Elo
67%
81.8%
84%
54.8 pts
76.4 pts
12%
64.3%
77.3%
84.4%
52.9%
67.8%
27.8%
39.3%
70.8%
8
GLM-5.2
93.3
62.1%
1584.3 Elo
44%
-
81%
-
-
13%
-
83%
84.2%
48.9%
62.4%
23.1%
34.3%
63.7%
9
Claude Sonnet 5
92.3
63.2%
1551.3 Elo
-
-
80.4%
-
-
-
-
-
-
-
-
-
-
-
10
GPT-5.6 Luna
90.4
62.7%
-
67.2%
77.9%
-
51.2 pts
74.6 pts
-
-
-
-
-
-
-
-
-
11
Qwen3.7-Max
90.2
60.6%
-
-
-
75%
-
-
-
-
78.3%
80.4%
47.2%
65.2%
18.4%
-
-
12
Muse Spark 1.1
89.6
61.5%
1535.8 Elo
53.3%
-
-
-
-
-
-
-
-
-
-
-
-
-
13
Claude Opus 4.6
89.4
57.3%
1542.1 Elo
-
66.6%
-
-
-
-
-
77.8%
80.8%
49.8%
70.8%
-
-
-
14
LongCat-2.0
88.4
59.5%
-
-
-
-
-
-
-
-
77.3%
-
-
-
-
-
-
15
混元3
87.0
57.9%
-
-
-
71.7%
-
-
-
-
75.8%
78%
45.6%
68.5%
23.8%
-
-
16
MiniMax M3
87.0
59%
1501.3 Elo
-
-
66%
-
-
-
-
-
80.5%
42.1%
74.5%
20.5%
37.1%
-
17
Claude Sonnet 4.6
86.7
58.1%
1521.5 Elo
-
-
67%
-
-
-
-
-
79.6%
-
68.3%
-
-
-
18
Kimi-K2.6
86.5
59.5%
1513.8 Elo
-
-
67.2%
-
-
-
-
77.1%
80.2%
42.8%
62.3%
-
-
48.3%
19
Seed2.1 Turbo
85.8
-
-
-
-
-
-
-
-
-
-
-
43.7%
-
-
-
49.4%
20
Gemini 3.5 Flash
85.7
-
1509.5 Elo
-
-
-
50.2 pts
-
-
-
-
-
-
-
-
-
-
21
GPT-5.4
85.3
57.7%
1456.9 Elo
-
79%
-
-
-
-
-
-
-
41.3%
60.3%
-
-
-
22
Kimi-K2.7-Code
85.3
-
1470.6 Elo
-
-
-
-
-
-
-
-
-
-
-
-
-
53.6%
23
Seed2.1 Pro
84.8
57.5%
1538.9 Elo
-
-
71%
-
-
-
-
75.1%
76%
47%
62.1%
23.5%
-
50.3%
24
Qwen3.7-Plus
84.4
57.6%
-
-
-
-
-
-
-
-
75.8%
77.7%
41.1%
-
-
-
-
低
高
单一有效来源基准按 0.5 权重计入,Arena.ai 除外;最高模型归一化为 100