Unverified50% confidenceBenchmarkExact time
SWE-Bench工程实战测试第一名是近期发布的MiniMax M2(397B)
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/10/2026
First Seen
Valid until: 10/8/2026
Sources
华为盘古2.0-Flash横评:指令遵循第一,代码工程垫底
bilibiliAI观模者6/30/2026
Related Claims
UnverifiedMiniMax M2.1在SWE-Bench基准测试上的表现持平甚至超过Claude Sonnet 4.575% similarUnverifiedMiniMax M2.7在SWE-Pro基准测试上得分56.22%73% similarUnverifiedQwen3.6-27B用27B参数量在SWE Bench Verified、SWE Bench Pro、Terminal Bench 2.0等多项基准测试中超越了前代397B MOE旗舰模型的性能70% similarUnverifiedMiniMax M3 的基准测试成绩亮眼,部分指标据称可比肩 GPT-5.567% similarUnverifiedMiniMax M3 在拖拽看板测试中得分60分,拖拽逻辑完整但缺乏国际化支持和响应式设计66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/463530API
curl https://kongchang.com/api/v1/knowledge/claims/463530MCP
get_claim(id=463530)