Lite
SWE-bench · 50 models
| Rank | Model | Score |
|---|---|---|
| 1 | ExpeRepair-v1.0 + Claude 4 Sonnet | 60.33 |
| 2 | Refact.ai Agent | 60 |
| 3 | KGCompass + Claude 4 Sonnet (20250514) | 58.33 |
| 4 | SWE-agent + Claude 4 Sonnet | 56.67 |
| 5 | Isoform | 55 |
| 6 | SemAgent_Multi-v1.0 | 51.67 |
| 7 | Isea | 51.33 |
| 8 | EntroPO + R2E + Qwen3-Coder-30B-A3B-Instruct | 49.67 |
| 9 | Blackbox AI Agent | 49 |
| 9 | Codev | 49 |
| 11 | Gru(2024-12-08) | 48.67 |
| 12 | ExpeRepair-v1.0 | 48.33 |
| 12 | Globant Code Fixer Agent | 48.33 |
| 14 | SWE-agent + Claude 3.7 Sonnet | 48 |
| 15 | devlo | 47.33 |
| 16 | DARS Agent | 47 |
| 17 | KGCompass + Claude 3.5 Sonnet (20241022) | 46 |
| 18 | EntroPO + R2E + Qwen3-Coder-30B-A3B-Instruct | 45 |
| 19 | Kodu-v1 + Claude-3.5 Sonnet (20241022) | 44.67 |
| 20 | CodeFuse-CGM | 44 |
| 21 | CodeStory Aide + Mixed Models | 43 |
| 22 | Lingxi | 42.67 |
| 23 | Codart AI | 41.67 |
| 23 | OpenHands + CodeAct v2.1 (claude-3-5-sonnet-20241022) | 41.67 |
| 25 | PatchKitty-0.9 + Claude-3.5 Sonnet (20241022) | 41.33 |
| 26 | Composio SWE-Kit (2024-10-30) | 41 |
| 26 | OrcaLoca + Agentless-1.5 + Claude-3.5 Sonnet (20241022) | 41 |
| 28 | Agentless-1.5 + Claude-3.5 Sonnet (20241022) | 40.67 |
| 29 | OpenCSG Starship Agentic Coder + GPT 4 (0806) | 39.67 |
| 30 | Bytedance MarsCode Agent | 39.33 |
| 31 | Moatless Tools + Claude 3.5 Sonnet (20241022) | 39 |
| 32 | Honeycomb | 38.33 |
| 32 | Moatless Tools + Claude 3.5 Sonnet (20241022) | 38.33 |
| 34 | AbanteAI MentatBot + GPT 4o (2024-05-13) | 38 |
| 35 | Patched.Codes Patchwork | 37 |
| 36 | KGCompass + DeepSeek V3 | 36.67 |
| 37 | AppMap Navie v2 | 36 |
| 38 | CodeFuse-AAIS | 35.67 |
| 38 | Gru(2024-08-11) | 35.67 |
| 40 | Isoform | 35 |
| 41 | Bytedance MarsCode Agent + GPT 4o (2024-05-13) | 34 |
| 41 | SuperCoder2.0 | 34 |
| 43 | Alibaba Lingma Agent | 33 |
| 44 | Agentless Lite + O3 Mini (20250214) | 32.33 |
| 45 | Agentless-1.5 + GPT 4o (2024-05-13) | 32 |
| 46 | CodeShellTester + GPT 4o (2024-05-13) | 31.33 |
| 46 | Factory Code Droid | 31.33 |
| 48 | AutoCodeRover (v20240620) + GPT 4o (2024-05-13) | 30.67 |
| 48 | Moatless Tools + Deepseek V3 | 30.67 |
| 50 | Aegis - o3-mini_1.0 | 30.33 |