Unverified80% confidenceFactTime unknown
在3D智慧城市监控大屏生成测试中，GLM-5.1获得9.8分，DeepSeek V4获得9.0分，GPT 5.5获得9.5分
1
Sources
80%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026
Sources
DeepSeek V4 vs GLM-5.1 vs GPT 5.5 编程实测：真实项目开发谁更强？
bilibili爱听书的程序员阿超
Related Entities
Related Claims
Unverified在3D跑酷游戏生成任务中，GPT 5.5两次生成均为空屏，获得0分，DeepSeek V4 Pro和MiniMax M3并列获得7分76% similarVerifiedDeepSeek-V3和DeepSeek-R1在多项基准测试中达到了与GPT-4o相当的水平72% similarUnverified在前端与视觉任务（Three.js动画、SVG图形）上，GPT-5.6模型普遍只能拿到六七分，而竞品Fable 5能稳定达到九到十分71% similarUnverifiedGPT-5.6编程能力在关键基准上达到91.9%，安全评测CyberGym得分85.6%，均超越竞品Nexus 571% similarUnverified在DeepSWE基准测试中，GPT 5.5以约70%的通过率高居榜首，比Opus 4.7领先超过15个百分点70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/25548API
curl https://kongchang.com/api/v1/knowledge/claims/25548MCP
get_claim(id=25548)