- Aggregate success rate rose from 19.6% to 41.2%, with gains spread across the whole task suite rather than one task
- Rope knotting and T-shirt unfolding were solved for the first time, after 143 prior failed T-shirt trials
- Evaluation ran six YAM workstations in parallel with 30 repeats per task and human-labeled failures, a fairly strict protocol
AI CURATED HOT STORY · 11/21/2026
GPT-6.2 Astra Doubles Embodied Task Success Rate to 41.2%
GPT-6.2 Astra has achieved a major breakthrough in robotic manipulation benchmarks, doubling its overall success rate from 19.6% (GPT-6 Astra two months ago) to 41.2%. In XDOF's Ember-0 dual-arm robot benchmark of 16 tasks, the improvement spans the entire task set rather than isolated successes. Notably, two flexible-object tasks—knotting a rope and unfolding a T-shirt—were solved by a model for the first time, with previous 143 T-shirt tests all failing.
Cross-validation reveals that GPT-6.2 Astra also excels in motion planning efficiency. On Artificial Analysis's AA-Robotics v0.1 benchmark, it issued only 668 commands across six tasks—the fewest among 11 models—while achieving the highest score. In the box-clearing task, it used just 92 arm commands, about 75% fewer than Claude Fable 5.2. XDOF's evaluation, using six YAM workstations in parallel with 30 repetitions per task and manual error annotation, is rigorous and reliable.
This progress signals a shift from single-point demos to general-purpose task capabilities in embodied AI. Flexible-object manipulation has long been a challenge; this breakthrough will accelerate deployment in homes and industries. Improved action efficiency also reduces deployment costs, paving the way for large-scale adoption. XDOF notes the benchmark may soon be saturated by model capabilities, implying faster-than-expected iteration.
Hourly heat curve
17 fully observed hoursThe line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.
Coverage timeline
Total 2 reports · Latest first- 量子位官网T2·78 pts
- Artificial Analysis (@ArtificialAnlys)T1·62 pts
- GPT-6.2 Astra issued 668 commands across six tasks, the lowest among 11 models and the highest score
- It used only 92 arm commands on box cleanup, about 75% fewer than Claude Fable 5.2
- AA-Robotics v0.1 measures robotics planning efficiency by command count, adding a new model comparison axis