OpenAI has published a post describing how the AI lab Third Moment Research runs 10,000 agents for more than 36 hours at a time. The company says the work involves building agents that can focus for days and improve with continued use.
According to the post, Third Moment Research relies on GPT-6 Astra and GPT-6.1 Sol, along with Codex and Daybreak for cybersecurity. These tools support the lab's agent systems.
The source text does not include benchmark results, pricing, or further technical details about how the agents are orchestrated or evaluated.