# Atomic Agent 发布了首个在基准测试中击败 Hermes 的本地代理，采用稳定前缀缓存和 TurboQuant 技术，在 GAIA Level 1 上完成 37 个任务

- 来源：Chubby
- 发布时间：2026-07-23 05:43
- AIWatch 分数：65
- AIWatch 标记：当日精选
- AIWatch 链接：https://aiwatch.icu/events/evt_01ky5z066ckkyjkrn1ta22aaxg
- 原文链接：https://x.com/kimmonismus/status/2080046135192777178

## 精选理由

常规快讯，保留列表

## AI 摘要

Atomic Agent 发布了首个在基准测试中击败 Hermes 的本地代理，采用稳定前缀缓存和 TurboQuant 技术，在 GAIA Level 1 上完成 37 个任务。

## 正文

• Runs Qwen, Gemma, and Llama via llama.cpp
• Stable-prefix caching keeps long sessions cheap
• TurboQuant shrinks the KV cache by 6.4×
• Solved 37 tasks vs. Hermes’ 31

I'm a huge fan of open source. Therefore, I'm delighted to see significantly more competition here. Local and open source is the future.

Atomic Agent: First local agent to beat Hermes on benchmarks!

✦ runs Qwen, Gemma, Llama via llama.cpp
✦ stable-prefix caching keeps sessions cheap
✦ TurboQuant cuts the KV-cache 6.4× smaller
✦ 37 tasks solved vs Hermes' 31 on GAIA Level 1

Open source on macOS, Windows & Linux 👇