# How Third Moment Research runs 10,000 agents for 36+ hours

Published: 2026-10-14 · Updated: 2026-10-14 · Category: Research

Source page: https://www.sipply.dev/story/how-third-moment-research-runs-10-000-agents-for-36-hours-c01fd867 · 中文: https://www.sipply.dev/zh/story/how-third-moment-research-runs-10-000-agents-for-36-hours-c01fd867

## Summary

Third Moment Research has made significant strides in the development of AI agents capable of operating continuously for over 36 hours. Utilizing advanced models like GPT-6 Astra and GPT-6.1 Sol, along with Codex and Daybreak, these agents are designed specifically for cybersecurity applications. Their ability to manage tasks effectively over extended periods showcases a remarkable improvement in performance, suggesting a new frontier in automated cybersecurity solutions.

## Key takeaways

- Third Moment Research has successfully developed AI agents that can run for over 36 hours.
- The agents utilize advanced models, indicating a leap in AI capabilities for task management.
- These developments are particularly relevant for the cybersecurity field, enhancing automated responses.

## Why it matters

The ability to run AI agents for extended periods represents a significant advancement in automation, particularly in critical fields like cybersecurity. This could lead to more robust and responsive systems that can adapt and improve over time.

- **Product managers:** For product managers, this development opens up new opportunities to integrate long-running AI agents into products, enhancing their functionality and reliability in real-world applications.
- **Developers:** Developers can leverage these advancements to create more efficient and resilient systems, particularly in cybersecurity, where continuous monitoring and response are crucial.
- **Students and job seekers:** For students and job-seekers, understanding these advancements in AI technology can provide insights into emerging career opportunities in AI development and cybersecurity.

## Sources

- [How Third Moment Research runs 10,000 agents for 36+ hours](https://openai.com/index/third-moment-research) (OpenAI News, 2026-10-14)

## 中文

### 第三时刻研究如何运行10,000个代理超过36小时

第三时刻研究在开发能够连续运行超过36小时的AI代理方面取得了重大进展。利用先进的模型如GPT-6 Astra和GPT-6.1 Sol，以及Codex和Daybreak，这些代理专门针对网络安全应用而设计。它们在长时间内有效管理任务的能力展示了性能的显著提升，暗示着自动化网络安全解决方案的新前沿。

**要点：**

- 第三时刻研究成功开发了可以运行超过36小时的AI代理。
- 这些代理利用先进的模型，表明任务管理的AI能力有了飞跃。
- 这些进展对网络安全领域尤其相关，增强了自动化响应能力。

**为什么重要：** 能够长时间运行AI代理代表了自动化的重大进步，特别是在网络安全等关键领域。这可能导致更强大和响应迅速的系统，能够随着时间的推移进行适应和改进。

---
Sipply: Sip AI News Simply. https://www.sipply.dev