← Arena leaderboards

Bash Reliability

Arena · 50 models

RankModelScore
1Claude Sonnet 5 (High)0.1498
2Claude Opus 5.5 (Max)0.1482
3GPT-6.1 Terra (Max)0.1367
4Claude Sonnet 5.5 (Max)0.1248
5GPT-6 Sol (Max)0.1136
6GPT-5.6 Sol (xHigh)0.1121
7Claude Fable 5 (High)0.1018
8Claude Opus 5 (Max)0.1006
9Gemini 3.8 Flash (High)0.0995
10Claude Fable 5.1 (Max)0.0993
11GPT-6 Luna (Max)0.0967
12Grok 4.6 (xHigh)0.0966
13Grok 4.7 (xHigh)0.0929
14GPT-6 Astra (Max)0.0839
15Muse Wonder Full (Max)0.0811
16Kimi K3.1 Pro (Max)0.0775
17Claude Haiku 5.5 (High)0.0739
18GPT-6 Luna (Medium)0.0684
19Gemini 3.7 Flash (High)0.0617
20GPT-6.1 Sol (Max)0.0514
21Muse Wonder (Max)0.0478
22Qwen3.7-Max0.0339
23DeepSeek-V4.1-Flash-20261017 (Max)0.033
24Inkling-Small0.0322
25Beam 1.0 501B A23B0.0313
26MiMo-V2.6-Pro0.0243
27Muse Spark 1.3 (Max)0.0186
28DeepSeek-V4-Pro (High, 20260813)0.0161
29GPT-6 Luna (Low)0.0144
30Inkling0.0108
31Gemini 3.1 Pro Preview0.0107
32Qwen3.7-Plus0.0073
33GPT-5.6 Terra (xHigh)0.0068
34Mistral Large 4-0.0069
35MiMo-V2.5-Pro-0.0095
36MiniMax-M3-0.0152
37Muse Spark 1.1-0.0257
38Qwen3.8-Max-0.0287
39Hy4 preview-0.0311
40Muse Spark 1.2 (xHigh)-0.0357
41MiMo-V2.6-Flash-0.0412
42Kimi K3 (Max)-0.0453
43GLM-5.2 (Max)-0.0456
44MiniMax-M3.1-Flash-0.0492
45GLM-5.3 (Max)-0.0542
46Gemini 4 Argon (High)-0.0555
47Gemini 4 Carbon (High)-0.0561
48Mistral Medium 3.5-0.079
49Spark-X2.6-0.0844
50GLM-5.3-Flash-0.1131