OpenAI Internal Data: Model Task Horizon Doubles Every 64 Days

OpenAI's official article reveals that the task time horizon its models can handle doubles every 64 days. Extrapolating from this, developer LeviTurk estimates a Bel loop agent achieves 41% success on 1024-hour tasks.

2026-10-17 ~ 2026-10-17 · 2 related posts

• OpenAI's internal data: model task time horizons double every 64 days — LeviTurk · 2026-10-17
• Extrapolating OpenAI's swarm scaling: Bel agents could run 1,024-hour tasks at 82% success — LeviTurk · 2026-10-17