Software Engineering on SWE-Bench Lite 300-issue subset
32AccuracyAgentless-1.5
Evaluation results
| Method | Links | |
|---|---|---|
| Agentless-1.5Open/Close Source=Open Source, Planning Model=GPT-4o2025.05 | 32 | |
| InfantAgent-NextOpen/Close Source=Open Source, Planning Model=GPT-4o2025.05 | 31.67 | |
| CodeShellTesterOpen/Close Source=Close Source, Planning Model=GPT-4o2025.05 | 31.33 | |
| Agentless + RepoGraphOpen/Close Source=Open Source, Planning Model=GPT-4o2025.05 | 29.67 | |
| SIMAOpen/Close Source=Close Source, Planning Model=GPT-4o2025.05 | 27.67 | |
| OpenHands + CodeAct v1.8Open/Close Source=Open Source, Planning Model=GPT-4o2025.05 | 22 |