← Leaderboard
RLDX-1
VLARLWRLD · May 5, 2026
VLA reported on RoboCasa365, LIBERO, LIBERO-Plus and SimplerEnv, with non-commercial weights. RLWRLD’s self-reported RoboCasa365 score (32.1) differs from the leaderboard value.
- Type
- VLA
- Parameters
- 6.9B
- Backbone
- Qwen3-VL-8B
- Weights
- Open
- License
- RLWRLD Model License v1.0
- Commercial use
- Not allowed
- Released
- May 5, 2026
- Robots
- Franka Panda (sim), WidowX (Bridge), Fourier GR-1 (sim), Franka Research 3, OpenArm, ALLEX
- SO-101
- –
- LeRobot
- –
Benchmark results
4 of 13 benchmarksRoboCasa365Simulation
- Official leaderboard ↗May 20, 2026
#17 of 2636.0%
LIBERO-PlusSimulation
- Paper / model card ↗May 2026
#6 of 2487.5%, self-reported by the authors, not a leaderboard entry
- Paper / model card ↗May 2026· n = 2000
#26 of 4697.8%, self-reported by the authors, not a leaderboard entry
- Paper / model card ↗May 2026· visual matching· n = 800
#9 of 1971.9%, self-reported by the authors, not a leaderboard entry
Not reported on 9 benchmarks
Robustness drop: 97.8% on LIBERO → 87.5% on LIBERO-Plus. The perturbed variants show how much of the LIBERO score is memorization.
Adoption
updated Nov 17, 2026Trained on
Runs on
Wrong or missing number? Report an error