Summary

OpenAI says its research organisation uses the equivalent of 3.1 agent-workdays for every human workday, with the median high-usage researcher spending more than $600 per day at API prices. The company also reports more code, more experiments and increasing delegation of higher-level tasks, while noting that human researchers still set priorities, judge results and intervene on complex work.

Why it matters

The disclosure is unusually concrete evidence about how coding agents are being deployed inside a frontier research organisation, while also showing that supervision and evaluation remain bottlenecks.

Sources