GPT-6 Astra on robot arms: 95% on block-in-bowl, still stuck on puzzle insertion

GPT-6 Astra on robot arms

Robocurve gave GPT-6 Astra control of YAM arms on two tasks, head-to-head with Claude Fable 5.1. On block-into-bowl, Astra scored 19/20 (95%) vs Fable 5.1's 8/20, averaging 2.5 min and $0.94 per run—less than half the time and cost of Fable 5.1's 6.8 min and $2.12. On the puzzle-insertion task, Astra managed 2/20, same as Fable 5.1; both stall at the final alignment step, at $1.36 per run. Clear win on pick-and-place, no progress on fine insertion.

Why it matters: Named first-person experiment with numbers and a direct model comparison — hits all three HKR axes. The puzzle-task stall for both models adds credibility. Not p1 because it's a third-party eval, not an official release, and only two tasks tested.