Logo

SWE Bench Leaderboard

Compare models on SWE Bench benchmarks.

RankModelScoreOrganizationLicense
1o3OpenAI69.1OpenAIProprietary
2o3OpenAI69.1OpenAIProprietary
3o4-miniOpenAI68.1%OpenAIProprietary
4Gemini 2.5 Pro PreviewGoogle63.8GoogleProprietary
5Claude 3.7 SonnetAnthropic62.3AnthropicProprietary
6GPT-4o miniOpenAI61%OpenAIProprietary
7o3-miniOpenAI61OpenAIProprietary
8GPT-4.1OpenAI55OpenAIProprietary
9Gemini 2.0 FlashGoogle51.8GoogleProprietary
10Gemini 2.0 FlashGoogle51.8%GoogleProprietary
11Claude 3.5 SonnetAnthropic49AnthropicProprietary
12o1OpenAI48.9OpenAIProprietary
13Claude 3.5 HaikuAnthropic40.6AnthropicProprietary
14Claude 3 HaikuAnthropic40.6AnthropicProprietary
15GPT-4o 2024-05-13OpenAI31OpenAIProprietary
16GPT-4oOpenAI31OpenAIProprietary
17GPT-4oOpenAI31OpenAIProprietary
18GPT-4oOpenAI31%OpenAIProprietary
19GPT-4.1 miniOpenAI23.6OpenAIProprietary