← Leaderboard

JoyAI-RA 0.5

World-action
JD Joy Future Academy · Aug 6, 2026

Later JoyAI-RA release, a vision-language-world-action model with a latent-action-conditioned world model. Results are self-reported: 92.0 average on seen tasks vs 74.0 for π0.5, and 75.5 on unseen tasks; no public-benchmark results.

Type
World-action
Parameters
—
Backbone
Vision-language-world-action model with latent-action world model + flow-matching action expert
Weights
Closed
License
–
Commercial use
Unclear
Released
Aug 6, 2026
Robots
AgiBot G1
SO-101
–
LeRobot
–
Inputs
–
VRAM
–

Benchmark results

0 of 13 benchmarks

No benchmark results recorded yet.

Not reported on 13 benchmarks
Wrong or missing number? Report an error