← Back to feed

GDPx agent benchmark: cost per point for 12 models, and the safety-filter caveat

AI · Studies · 2026-11-05 19:13

SemiAnalysis发布GDPx基准，涵盖206项长程任务，最高分仅12%，文章解析各模型每点成本及安全过滤限制。

Source: theainewsreport.com

Auto-generated summary for reference only. RSS