返回探索

实测 JEV 路由 AI 代理:换种提问方式后追平大模型,延迟仅为三分之一

作者花了一天时间测试 JEV 是否能比主流大模型更好地路由 AI 代理。初期结果不如对方,但改变提问方式后打成平手,延迟降至三分之一,成本也大幅降低。帖中还分享了完整测量数据和失败场景。

I spent a day trying to prove JEV could route my AI agents better than the big ones. It lost. Then I changed how I asked the question and it tied them, at a third of the latency and a fraction of the cost. Everything I measured, including where it fails 🧵

实测 JEV 路由 AI 代理:换种提问方式后追平大模型,延迟仅为三分之一 1
· 1 次赞在 X 打开