The Decoder· Manuel Uth·· 3 天前AI 评分49
No AI agent scores above 12 percent on SemiAnalysis' new real-world work benchmark
No AI agent scores above 12 percent on SemiAnalysis' new real-world work benchmark
AI 导读
SemiAnalysis 发布 GDPx benchmark，将 AI agents 置于模拟知识工作环境，无模型得分超过 12 percent。Claude Opus 5.5 排名第一，得分为 12 percent；Gemini 4 Carbon 得分 10.9 percent，成本为 $16.29 per task。
来源：The Decoder · the-decoder.com