AgentTime Benchmark: GPT-6 Astra Hits 63% On-Time Rate in 72-Hour Tasks

A new AgentTime benchmark (arXiv:2610.09944) systematically tests whether AI agents can work for requested durations, predict runtime, and estimate elapsed time. Across 1,991 runs, GPT-6 Astra achieved a 63% on-time rate on 72-hour tasks, far outperforming rivals like Fable 5.1, which quit after three hours.

2026-10-12 ~ 2026-10-12 · 2 related posts

• GPT-6 Astra nails 72-hour agent tasks with 63% on-time rate, crushing Fable 5.1's 4% — maksym_andr · 2026-10-12
• AgentTime benchmark: GPT-6 Astra nails a 72-hour task while Fable 5.1 quits after 3 hours — maksym_andr · 2026-10-12