Multi-SWE-Bench (c++): leaderboard
Metric: Score (%). Source: huggingface.co. 38 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | CodeArts Agent + CodeArts-GLM-5.1 | 47.29 |
| 2 | InfCode + GPT5(Sep) | 25.58 |
| 3 | MopenHands + Claude-3.7-Sonnet | 14.73 |
| 4 | MopenHands + Claude-3.5-Sonnet(Oct) | 12.4 |
| 5 | MSWE-agent + Claude-3.7-Sonnet | 11.63 |
| 6 | MopenHands + Gemini-2.5-Pro | 9.3 |
| 7 | MSWE-agent + Gemini-2.5-Pro | 8.53 |
| 8 | MSWE-agent + DeepSeek-V3 | 7.75 |
| 9 | MopenHands + DeepSeek-V3 | 7.75 |
| 10 | MSWE-agent + Claude-3.5-Sonnet(Oct) | 6.98 |
Interactive version: theaggregate.ai/benchmark?slug=multi-swe-bench-c-plus-plus · How It Works · Data refreshed daily, snapshot 2026-09-05.