GPT-6 Astra's Robotics Performance on RoboDojo and Cost Comparison with Grok 4.7
A research paper, likely from the developers, presented early evaluations of GPT-6 Astra's robotics policy on RoboDojo. Concurrently, MIXED Reality News reported that Grok 4.7 achieved performance comparable to GPT-6 Astra's cheapest setting, but at a significantly higher cost per task. The primary claims regarding GPT-6 Astra's capabilities are detailed in the first-party research.
2 articles from 2 outlets covered this story. The underlying claim is sourced from a paper.
What do all outlets agree on?
2 outlets covered “GPT-6 Astra's Robotics Performance on RoboDojo and Cost Comparison with Grok 4.7”. All of them report the following:
- GPT-6 Astra's robotics policy was evaluated
- Evaluations were conducted on RoboDojo
- Grok 4.7 was compared to GPT-6 Astra's performance
- Grok 4.7 achieved performance comparable to GPT-6 Astra's cheapest setting
Which outlets covered this?
All 2 articles found on this story, grouped by the stance of the piece. Every link goes to the original publisher.
pragmatist2