Unverified50% confidenceBenchmarkExact time
MiniCPM5-2B在SWE-bench上的成绩击败了参数量是其两倍的4B级对手
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/16/2026
First Seen
Valid until: 12/15/2026
Sources
Related Entities
Related Claims
UnverifiedMiniCPM5-2B在SWE-bench上取得46.4分76% similarUnverifiedMiniCPM5-2B在Artificial Analysis Intelligence Index v4.2基准测试中获得15分综合成绩,在4B参数以下开源模型中排名第一69% similarUnverifiedMiniCPM 系列以 2B-4B 级别的参数量在多项基准测试中实现了接近 7B 甚至 13B 模型的性能69% similarUnverified在BU Bench V1上,4B参数的SFT模型超越了DeepSeek V4 Pro和Kimi K2.6在开源Browser Use框架下报告的45%成绩65% 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64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/924466API
curl https://kongchang.com/api/v1/knowledge/claims/924466MCP
get_claim(id=924466)