Robot Manipulation (Grasp and Success Rates) on SimplerEnv WidowX
62.5Success Rate: Put SpoonLAST-Continuous
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| LAST-ContinuousAction Generation Mode=Continuous2026.05 | 62.5 | 87.5 | 79.2 | 41.7 | 79.2 | 33.3 | 100 | 91.7 | — | 57.3 | |
| LAST-DiscretizedAction Generation Mode=Discretized2026.05 | 58.3 | 83.3 | 79.2 | 37.5 | 83.3 | 29.2 | 100 | 95.8 | — | 55.2 | |
| Octo-SmallModel Scale=Small2025.11 | 47.2 | 77.8 | 27.8 | 9.7 | 40.3 | 4.2 | 87.5 | 56.9 | 58.4 | 29.5 | |
| Octo-SmallModel Scale=Small2026.05 | 47.2 | 77.8 | 27.8 | 9.7 | 40.3 | 4.2 | 87.5 | 56.9 | — | 29.5 | |
| RoboVLMs2026.05 | 45.8 | 70.8 | 33.3 | 20.8 | 54.2 | 4.2 | 91.7 | 79.2 | — | 37.5 | |
| π0-FAST-ContinuousModel Variant=FAST, Action Generation Mode=Continuous2026.05 | 45.8 | 70.8 | 66.7 | 41.7 | 58.3 | 25 | 79.2 | 70.8 | — | 45.8 | |
| BEAST2026.05 | 41.7 | 66.7 | 37.5 | 25 | 50 | 20.8 | 87.5 | 75 | — | 37.5 | |
| BEAST-ContinuousAction Generation Mode=Continuous2026.05 | 37.5 | 54.2 | 70.8 | 41.7 | 33.3 | 12.5 | 87.5 | 70.8 | — | 40.6 | |
| OpenVLA-OFTModel Variant=OFT2026.05 | 34.5 | — | — | 30 | — | 30 | — | 72.5 | — | 41.8 | |
| PixelVLA-π0Base Architecture=π02025.11 | 32.4 | 51.7 | 28.7 | 16.7 | 56.8 | 21.7 | 83.3 | 61.7 | 55.1 | 33.8 | |
| π0-FASTModel Variant=FAST2026.05 | 29.2 | 33.3 | 25 | 20.8 | 37.5 | 12.5 | 75 | 66.7 | — | 32.3 | |
| π0Zero-shot Evaluation=true2025.11 | 29.1 | 45.8 | 25 | 0 | 50 | 16.6 | 91.6 | 62.5 | 53.1 | 27.1 | |
| π02026.05 | 29.1 | — | — | 0 | — | 16.6 | — | 62.5 | — | 27.1 | |
| π0-SFTSupervised Fine-Tuning (SFT)=true2025.11 | 26.8 | 45.3 | 28.6 | 4.2 | 52.3 | 18.6 | 88.5 | 59.6 | 53.7 | 27.3 | |
| RoboVLMs2025.11 | 20.8 | 37.5 | 33.3 | 25 | 8.3 | 8.3 | 0 | 0 | 19.8 | 13.5 | |
| SpatialVLA2025.11 | 20.8 | 25 | 41.7 | 20.8 | 58.3 | 25 | 79.2 | 70.8 | 51.1 | 34.4 | |
| SpatialVLA2026.05 | 16.7 | 20.8 | 29.2 | 25 | 62.5 | 29.2 | 100 | 100 | — | 42.7 | |
| Octo-BaseModel Scale=Base2025.11 | 12.5 | 34.7 | 52.8 | 8.3 | 31.9 | 0 | 66.7 | 43.1 | 46.5 | 16 | |
| Octo-BaseModel Scale=Base2026.05 | 12.5 | 34.7 | 52.8 | 8.3 | 31.9 | 0 | 66.7 | 43.1 | — | 16 | |
| PixelVLABase Architecture=OpenVLA2025.11 | 4.2 | 20.8 | 37.5 | 20.8 | 16.6 | 0 | 79.2 | 41.7 | 38.5 | 16.7 | |
| RT-1-X2025.11 | 0 | 16.7 | 20.8 | 4.2 | 8.3 | 0 | 0 | 0 | 11.5 | 1.1 | |
| OpenVLAZero-shot Evaluation=true2025.11 | 0 | 4.1 | 33.3 | 0 | 12.5 | 0 | 8.3 | 4.1 | 14.6 | 1 | |
| OpenVLA-SFTSupervised Fine-Tuning (SFT)=true2025.11 | 0 | 8.4 | 35.1 | 12.8 | 10.5 | 0 | 16.5 | 8.4 | 17.6 | 5.3 |