VLA-Arena Overall: leaderboard
Vision-language-action policies across all 11 VLA-Arena manipulation tasks, averaged over three difficulty levels covering safety, distractors, extrapolation and long-horizon control.
Metric: Success Rate (%). Source: vla-arena.github.io. 30 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Paimon | 73.08 | |
| 2 | MolmoAct2 | 72.12 | |
| 3 | StellaVLA | 68.12 | |
| 4 | ACoT-VLA | 66.09 | |
| 5 | DM0.5 | 62.26 | |
| 6 | LingBot-VLA 2.0 | 54.46 | |
| 7 | GR00T-N1.7 | 52.49 | |
| 8 | X-Policy | 51.24 | |
| 9 | LingBot-VLA | 50.93 | |
| 10 | LangForce | 49.03 |
Interactive version: theaggregate.ai/benchmark?slug=vla-arena-overall · How It Works · Data refreshed daily, snapshot 2026-10-20.