Event ·Multi-source

Agent Arena 排行:复杂任务上 Gemini 4 Argon 与小米 MiMo 表现突出

Agent Arena 数据显示,在低成本任务上 OpenAI 和 Anthropic 占据 11 个前沿席位中的 10 个,但在最昂贵的 5% 会话中只剩 7 个。空出的位置由 Google 的 Gemini 4 Argon 和小米的 MiMo-V2.6 Flash 与 Pro 补上。近两周有超过 1,000 个会话成本超过 100 美元,选错模型可能明显推高账单。五大实验室旗舰模型对比中,Claude Opus 5.5 在长时程任务上耗时最长,峰值达 1 小时 51 分钟,Kimi K3.1 Pro 耗时最短,为 56 分钟,两者相差 55 分钟。

Current read

Agent Arena 拆了成本维度看模型:贵任务上 OpenAI 和 Anthropic 不再通吃,Gemini 4 Argon 和小米 MiMo 顶上,选型前值得对号入座。

11 sources39° AI heatLast updated 2026/11/4 17:41:21

Evidence

11 sources
Related source 2rohanpaul_ai
View source
Related source 4Shubham Saboo
View source
Related source 5TestingCatalog
View source
Related source 6Nathan Lambert
View source
Related source 7Simon Willison’s Weblog
View source
Related source 8Claude Code: GitHub Releases
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.