Primary sourceBindu Reddy
View sourceEvent ·Multi-source
Gemini Argon 4 在 LiveBench 2.0 上落后 Fable 与 Astra，仍领先开源模型
Gemini Argon 4.0 在自建的 LiveBench 2.0 基准上落后于 GPT 6.1 Sol，也排在前面的 Fable 和 Astra 之后。该模型仍领先开源的 DeepSeek Flash，并明显优于 Muse Wonder，而 Muse 只比 DeepSeek 略好。发布方称 LiveBench 2.0 难以刷分，更接近真实使用场景。
Current read
用 LiveBench 2.0 给 Gemini Argon 4、GPT 6.1 Sol 和 DeepSeek Flash 排了个序，榜单自称难刷分，可以看看结果信不信得过。
11 sources57° AI heatLast updated 2026/10/13 16:01:39
Evidence
11 sourcesRelated source 1Blognone
View sourceRelated source 2TestingCatalog
View sourceRelated source 3Epoch AI
View sourceRelated source 4Artificial Analysis
View sourceRelated source 5IT之家
View sourceRelated source 6lmarena.ai
View sourceRelated source 7kimmonismus
View sourceRelated source 8The Rundown AI
View sourceRelated source 9Ethan Mollick
View sourceRelated source 10Alexandr Wang (Meta MSL)
View sourceConflict 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.