MMODELYST
Leaderboards/SWE-bench Verified Leaderboard
LDR

SWE-bench Verified Leaderboard

SWE-bench Verified

Real-world software-engineering issue resolution ranking.

1Claude 3.7 Sonnet
62.3
2GPT-4.1
54.6
3Claude 3.5 Sonnet
49
4o1
48.9
5Claude 3.5 Haiku
40.6