# Apodex-1.0-H 发布多智能体深度研究团队

- 来源：Rohan Paul (@rohanpaul_ai)
- 发布时间：2026-06-11 01:07
- AIHOT 分数：64
- AIHOT 链接：https://aihot.virxact.com/items/cmq8c7gms0135slldatzl87tt
- 原文链接：https://x.com/rohanpaul_ai/status/2064756356805259323

## AI 摘要

Apodex-1.0-H 发布一个异步智能体团队，用于深度研究。协调者将子智能体分配到独立上下文和工具，再通过事实核查、冲突审查和草稿审查智能体检验弱主张。该方案将深度研究视为分布式系统问题，展示了推理时缩放路径：通过多个协调搜索智能体、持久追踪和独立验证层提升答案质量，而非依赖单一更大模型，并声称取得 SOTA 结果。

## 正文

Apodex-1.0-H just dropped a heavy-duty agent team for deep research

Claims SOTA results by splitting web research across many agents and auditing every evidence chain before writing the answer.

Treats deep research as a distributed systems problem for AI agents.

Apodex uses an async agent team： an orchestrator assigns sub-agents separate contexts and tools， then fact-checker， conflict-reviewer， and draft-reviewer agents test weak claims.

The real big deal is that Apodex is showing a possible "inference-time scaling" path for AI research， where better answers come not from one bigger model， but from many coordinated search agents， persistent traces， and a separate verification layer that audits the evidence before the final response is allowed to exist.

### 引用推文

> Apodex：Dive in 👇 📝 Blog: https://www.apodex.com/blog/apodex-1.0 📄 Tech report: http://www.apodex.com/pdf/20260608 💻 Github: https://github.com/ApodexAI/AgentHarnes...