← Back to events
ActiveAI

GPT-6 Astra on robot arms

Photo: Hacker News

What happened

OpenAI's GPT-6 Astra vs Claude Fable 5.1 controlling a pair of YAM arms under the same agent policy: 19/20 vs 8/20 on block-into-bowl in interleaved blinded pairs, 2/20 vs 2/20 on the puzzle, with 80% fewer output tokens.

Summary assembled by rule from the sources below

Why it's spreading

Sources

Community