Unverified60% confidenceFactTime unknown
GLM-4.7是一款MOE（混合专家）架构模型，总参数量达358B（3580亿）
2
Sources
60%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
GLM-4.7深度实测：编程能力全面对标Claude Sonnet 4.5
bilibiliAI超元域
Related Entities
Related Claims
VerifiedGLM-4.6采用专家混合架构（MoE），总参数量达到3550亿，推理时只激活约320亿个参数86% similarUnverified智谱GLM系列从GLM4.5到4.7在358B参数规模上连续打磨三代，随后GLM5翻倍至754B78% similarUnverifiedGLM5的参数量约为745B，激活参数约为44B，参数激活比约为1:1774% similarUnverifiedGLM4在World of AI基准测试工具中排名第五73% similarUnverifiedGLM 4.6在HumanEval、MBPP、SWE-bench等代码评测基准以及MMLU、GPQA等综合能力基准上的表现对齐Claude 3.4和Claude 3.4.572% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/24226API
curl https://kongchang.com/api/v1/knowledge/claims/24226MCP
get_claim(id=24226)