Unverified90% confidenceFactTime unknown
在简单题中，GLM 5.1补了19行测试，Kimi K2.6补了12行，其他五个模型（包括Sonnet 4.6）一行测试都没写
1
Sources
90%
Confidence
Medium-term (~90 days)
Relevance
5/31/2026
First Seen
Valid until: 8/29/2026
Sources
7个AI修同一个Bug实测：GLM 5.1反超Claude Sonnet详细对比
bilibili寒冰巨魔9527
Related Entities
Related Claims
Unverified在普通题（MCP协议客户端Signal转发）中，GLM 5.1写了108行测试，是Sonnet的1.3倍，Kimi的2倍75% similarUnverifiedGLM4在World of AI基准测试工具中排名第五67% similarUnverified国产模型中编程能力最强的GLM 5.1，实测水平大致处于Claude Opus 4.5和Sonnet 4.6之间67% similarUnverifiedGLM-4.7在代码能力基准测试中超过DeepSeek V3.2及Claude Sonnet 4.567% similarUnverified在图片隐藏虫子识别测试中，Sonnet 5的Max模式花费约10分钟并识别结果完全正确，而Gemini顶配模型只找到4只，GPT-5.5 X-High未能检测出66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/13174API
curl https://kongchang.com/api/v1/knowledge/claims/13174MCP
get_claim(id=13174)