How Third Moment Research runs 10,000 agents for 36+ hours
- ID
- 35383
- Status
- summarized
- Published
- 14 Oct 2026, 3:00 PM
- Fetched
- 15 Oct 2026, 12:08 AM
- Provider
- OpenAI News
- Category
- ai-labs
- Original URL
- https://openai.com/index/third-moment-research
- Source URL
- https://openai.com/news/rss.xml
Summary
- Score
- 6.0
- Created
- 15 Oct 2026, 12:08 AM
- Tags
- Audience
- developersai_agent_usersai_ml_learnerssaas_startup_founders
What happened
Third Moment Research uses OpenAI's GPT-6 Astra, GPT-6 Sol, GPT-6.1 Sol, Codex, and Daybreak to run 10,000 concurrent agents in 36–48 hour autonomous research sessions, with a self-improving loop that generates thousands of synthetic scenarios around recurring agent weaknesses. In an early quantitative-finance use case, Codex analyzed thousands of simulation trajectories and found agents were wasting tokens polling for backtests to finish; after a caching and memory-packing fix, a year-long order-level US equity backtest dropped from about 3 hours to about 3–5 minutes. The article also cites 1 trillion OpenAI tokens used in a single day and a 97% reduction in backtest time with Codex.
Why it matters
For builders running long-lived agents, the concrete failure mode is status polling filling context and burning tokens; the fix here combined caching and memory packing and cut backtest time from ~3 hours to ~3–5 minutes. Malaysian agent builders can apply that pattern to reduce API cost and context pollution, but the article is an OpenAI customer story with no code, independent benchmark, or Malaysia-specific policy/funding angle.
Discussion angle
What would it take to instrument your own agent loop for token-wasting status polling, and would a caching/memory-packing fix be worth the engineering time before you scale to hundreds or thousands of concurrent agents?