Unverified50% confidenceFactExact time
GLM-5在SWE-Bench Verified基准测试中得分77.8%，位于开源第一梯队
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026
Sources
GLM-5.2深度解析：百万上下文、MIT开源与国产算力全栈突破
bilibiliAI技术投降派6/14/2026
Related Claims
UnverifiedGLM4在SWE-Bench基准测试中得分46.2%，显著优于上一代模型79% similarUnverifiedGLM 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78% similarUnverifiedGLM4在World of AI基准测试工具中排名第五78% 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74% similarUnverifiedGLM 5.2的Diff测试分数从GLM 5.1的18分跃升至46.2分73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/46081API
curl https://kongchang.com/api/v1/knowledge/claims/46081MCP
get_claim(id=46081)