OPENAI · 2026-10-14 15:01 UTC

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

Third Moment Research runs 10,000 AI agents simultaneously for 36-48 hour investigations using OpenAI's latest models (GPT-6 Astra, GPT-6 Sol, GPT-6.1 Sol) to build self-improving systems that identify weaknesses and create synthetic scenarios for improvement. Using Codex, the company optimized a backtesting bottleneck in quantitative finance from 3 hours to 3-5 minutes, reducing token waste and context pollution for agents. Agents work within a Dockerized environment with terabytes of data and tools to improve baseline trading systems by testing strategies, assessing results, and…

Loading the story on Tree News…