Event ·Multi-source

Artificial Analysis 发布 Intelligence Index v5,新增两个基准并引入私有测试集

Artificial Analysis 发布 Intelligence Index v5,这是自 1 月 v4 之后最大一次升级,包含 2 个新基准和 3 项更新,60% 的指数权重基于私有测试集或保留答案的评测。新增基准包括 Cognition 的 FrontierCode 2.0 和 70 个终端环境下智能体科研任务的 Terminal-Bench-Science。榜单上 Claude Opus 5.5 (max) 以 58 分居首,GPT-6 Astra (max) 以 57 分紧随其后。开源权重方面,Kimi K3.1 Pro 以 46 分最高,已发布权重中 DeepSeek V4.1 Flash 1017 (max) 与 GLM-5.3 (max) 均为 32 分,与榜首差距从 3 分扩大到 12 分。同时发布的 Progress Index v1 用于在同一尺度上追踪模型能力随时间的变化。

Current read

模型评测榜 v5 版来了,60% 改用私有题防刷榜,Claude Opus 5.5 拿了第一,开源和闭源差距拉大到 12 分,选模型前可以看看。

11 sources74° AI heatLast updated 2026/10/28 02:11:22

Evidence

11 sources
Primary sourceArtificial Analysis
View source
Related source 1kimmonismus
View source
Related source 2arXiv: DeepSeek
View source
Related source 3The Rundown AI
View source
Related source 5TestingCatalog
View source
Related source 6Ethan Mollick
View source
Related source 8Simon Willison’s Weblog
View source
Related source 9marktechpost
View source
Related source 10量子位
View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.