Unverified50% confidenceFactExact time
GLM 5.2的Diff测试分数从GLM 5.1的18分跃升至46.2分
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026
Sources
GLM 5.2与Zcode实测：每天500万Token免费AI编程工具深度体验
bilibiliNixAI6/18/2026
Related Claims
UnverifiedGLM 5.2在Terminal Bench测试中得分81分，GLM 5.1得分63.5分，GPT 5.5得分84分，Opus 4.8得分85分83% 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754B77% similarUnverifiedGLM 5.2在Artificial Analysis Intelligence Index上，MAX版本得分51分，是目前得分最高的开源权重模型75% similarUnverifiedGLM 5.2在Frontier SW基准测试中得分为74.4%，接近Anthropic Opus 4.8的水平75% similarUnverifiedGLM4在SWE-Bench基准测试中得分46.2%，显著优于上一代模型74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/43920API
curl https://kongchang.com/api/v1/knowledge/claims/43920MCP
get_claim(id=43920)