Skip to content

Back to the catalog

openai.comarticle

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

OpenAI customer case study on Third Moment Research, which runs up to 10,000 concurrent agents on 36–48 hour autonomous quantitative-finance research sessions. Agents use the default Codex harness in a Dockerized environment with a terabyte of data. An outer harness driven by GPT-6 Sol or GPT-6.1 Sol watches for unfinished tasks and decides whether to retry, terminate or escalate to a human. Codex analyzed thousands of recorded trajectories and found a backtest bottleneck. A caching and memory-packing fix cut the backtest from about 3 hours to 3–5 minutes, which also reduced agent polling and context pollution. The post also covers how the team picks models by token efficiency and speedup benchmarks (GPT-6.1 Sol, GPT-6 Astra, GPT-5.6 Terra), and says Daybreak automates penetration tests of the sandbox.

TL;DR Third Moment Research runs 10,000 Codex-harness agents for 36–48 hours, with an outer GPT-6.x supervisor harness and trajectory analysis that cut a backtest from 3 hours to minutes.
Open original ↗

Organisations named here

Technologies named here

Techniques named here

  • Self-improving agent loop with synthetic scenarios
  • Trajectory analysis with Codex
  • Outer harness supervising Codex as subagent
  • Parallel subagent root-cause investigation
  • Automated agent-driven penetration testing of environments

Related in agents