BenchMark·Hub

SWE-bench Bash-only

A已审核Variant
代码代码生成SWE-bench 家族未知

发布方:Princeton

热度7.1▲ 0.1

SWE-bench Bash-only

swe-bench / resolved · 47 行结果

排名模型子项分数Metric时间
#1Claude 4.5 Opus
AnthropicClaude 4.5 Opus (high)
SWE-bench Bash-only
swe-bench
76.8%resolved2026-02-17
#2Gemini 3 Flash
Google DeepMindGemini 3 Flash (high)
SWE-bench Bash-only
swe-bench
75.8%resolved2026-02-17
#3MiniMax M2.5
MinimaxMiniMax M2.5 (high)
SWE-bench Bash-only
swe-bench
75.8%resolved2026-02-17
#4Claude 4.6 Opus
Anthropic
SWE-bench Bash-only
swe-bench
75.6%resolved2026-02-17
#5Claude 4.5 Opus
AnthropicClaude 4.5 Opus (20251101) (medium)
SWE-bench Bash-only
swe-bench
74.4%resolved2025-11-24
#6Gemini 3 Pro Preview
Google DeepMindGemini 3 Pro Preview (2025-11-18)
SWE-bench Bash-only
swe-bench
74.2%resolved2025-11-18
#7GPT 5.2 Codex
OpenAI
SWE-bench Bash-only
swe-bench
72.8%resolved2026-02-19
#8GLM 5
Z-AIGLM 5 (high)
SWE-bench Bash-only
swe-bench
72.8%resolved2026-02-17
#9GPT 5.2
OpenAIGPT 5.2 (high)
SWE-bench Bash-only
swe-bench
72.8%resolved2026-02-17
#10GPT 5.2
OpenAIGPT 5.2 (2025-12-11) (high)
SWE-bench Bash-only
swe-bench
71.8%resolved2025-12-11
#11Claude 4.5 Sonnet
AnthropicClaude 4.5 Sonnet (high)
SWE-bench Bash-only
swe-bench
71.4%resolved2026-02-17
#12Kimi K2.5
Moonshot AIKimi K2.5 (high)
SWE-bench Bash-only
swe-bench
70.8%resolved2026-02-17
#13Claude 4.5 Sonnet
AnthropicClaude 4.5 Sonnet (20250929)
SWE-bench Bash-only
swe-bench
70.6%resolved2025-09-29
#14DeepSeek V3.2
DeepSeekDeepSeek V3.2 (high)
SWE-bench Bash-only
swe-bench
70%resolved2026-02-17
#15Gemini 3 Pro
Google DeepMind
SWE-bench Bash-only
swe-bench
69.6%resolved2026-02-26
#16GPT 5.2
OpenAIGPT 5.2 (2025-12-11)
SWE-bench Bash-only
swe-bench
69%resolved2025-12-11
#17Claude 4 Opus
AnthropicClaude 4 Opus (20250514)
SWE-bench Bash-only
swe-bench
67.6%resolved2025-08-02
#18Claude 4.5 Haiku
AnthropicClaude 4.5 Haiku (high)
SWE-bench Bash-only
swe-bench
66.6%resolved2026-02-17
#19GPT 5.1 Codex
OpenAIGPT 5.1 Codex (medium)
SWE-bench Bash-only
swe-bench
66%resolved2025-11-24
#20GPT 5.1
OpenAIGPT 5.1 (2025-11-13) (medium)
SWE-bench Bash-only
swe-bench
66%resolved2025-11-20
#21GPT 5
OpenAIGPT 5 (2025-08-07) (medium)
SWE-bench Bash-only
swe-bench
65%resolved2025-08-07
#22Claude 4 Sonnet
AnthropicClaude 4 Sonnet (20250514)
SWE-bench Bash-only
swe-bench
64.9%resolved2025-07-26
#23Kimi K2 Thinking
Moonshot AI
SWE-bench Bash-only
swe-bench
63.4%resolved2025-12-10
#24MiniMax M2
Minimax
SWE-bench Bash-only
swe-bench
61%resolved2025-11-24
#25DeepSeek V3.2 Reasoner
deepseek
SWE-bench Bash-only
swe-bench
60%resolved2025-12-01
47 条结果 · 当前 1-25
1 / 2