Artificial Analysis 改版智能指数,开闭源模型差距被拉大
The Decoder · rss · 2026-10-28
The Decoder reports that Artificial Analysis overhauled its Intelligence Index, and under the new methodology the gap between open and closed models widens noticeably. Kimi K3.1 Pro dropped five spots in the rankings. The revamp has sparked discussion about how benchmark methodology changes affect the relative standing of open-weight versus closed models.
「模型」频道最新
- 医疗公司运营部门单日烧掉11.9亿枚 Opus token — RishabJainK · 2026-10-30
- 曝 Kimi K3.1 Pro 系列将亮相，11 月 4 日开放权重，_progress index_ 已超 Fable 5 — teortaxesTex · 2026-10-30
- 用户吐槽 Opus 5.5 输出「难以卒读」，要求改写后仍是技术腔 — msalbergo · 2026-10-30
- 科研智能体用模型调研:DeepSeek V4.1 Flash 成最热门开源模型 — AkariAsai · 2026-10-30
- 圈内人称 Gemini Argon 会做出研究者级「跳跃」，Google 疑似掌握独家诀窍 — vitaliychiley · 2026-10-30
- Alexandr Wang 晒图:Meta Muse 回答比 ChatGPT 简洁 6.7 倍 — alexandr_wang · 2026-10-30