返回
原文
返回
AI 评分 41/100在 X 查看原推
导出 Markdown

Atomic Agent 成为首个在 GAIA 基准上超越 Hermes 的开源本地 AI 智能体

Rohan Paul · @rohanpaul_ai · X·2026-07-23 07:11·36天前
AI 导读

Atomic Agent 成为首个在 GAIA 基准上超越 Hermes 的开源本地 AI 智能体。使用 Qwen-3.6-35b-A3B 模型，在相同预算下解决 37/53 个任务（Hermes 为 31 个），且运行更快。其 TurboQuant 技术将 KV-cache 缩小 6.4 倍并提速 30-50%，使 9B 模型也能完成 52.8% 的任务。

Rohan Paul@rohanpaul_ai
41AI 编辑部评分，满分 100

Atomic Agent 成为首个在 GAIA 基准上超越 Hermes 的开源本地 AI 智能体

2026-07-23 07:11· 36天前
AI 导读

Atomic Agent 成为首个在 GAIA 基准上超越 Hermes 的开源本地 AI 智能体。使用 Qwen-3.6-35b-A3B 模型，在相同预算下解决 37/53 个任务（Hermes 为 31 个），且运行更快。其 TurboQuant 技术将 KV-cache 缩小 6.4 倍并提速 30-50%，使 9B 模型也能完成 52.8% 的任务。

The 1st open-source local agent to surpass Hermes on GAIA benchmark.

Loop design, not model size, drove this whole gap.

@atomicagent_io , a local-first AI agent that runs on your machine, with local or cloud models. Open source on macOS, Windows & Linux.

Both agents drove qwen-3.6-35b-a3b, on the same task budget. Atomic Agent solved 37 of 53 tasks, against 31 for Hermes, and it ran faster too.

Most agent loops get costly because every turn re-sends the whole growing context, and small local models choke on that first.

Atomic Agent keeps turns cheap. One inference returns a grammar-checked JSON array of tool calls, reads run in parallel, results get summarized, and the fixed prompt reuses cached work.

Underneath sits a custom llama.cpp build called TurboQuant. It shrinks the memory cache up to 6.4 times and adds 30-50% more speed through tailored decoding. This is how a 9B model, qwen-3.5-9b, still cleared over half the same tasks at 52.8%.

Atomic AgentFirst local agent to beat Hermes on benchmarks! ✦ runs Qwen, Gemma, Llama via llama.cpp ✦ stable-prefix caching keeps sessions cheap ✦ TurboQuant cuts the KV-ca...

来源：Rohan Paul· x.com

智能体产品更新开源/仓库评测/基准
导出 Markdown