RoboChallenge Table 30: leaderboard
Real-robot evaluation on 30 tabletop tasks executed on physical arms rather than in simulation, covering object reorientation, insertion, food preparation and multi-stage tool use.
Metric: Score (0-100). Source: robochallenge.ai. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Era0 | 76.34 | |
| 2 | DM0 | 72.25 | |
| 3 | GigaBrain-0.1 | 68.34 | |
| 4 | Spirit-v1.5 | 67.19 | |
| 5 | pi0.5 | 61.84 | |
| 6 | Lira_generalist | 59.83 | |
| 7 | wall-oss-v0.1 | 55.3 | |
| 8 | Atlas_generalist | 55.1 | |
| 9 | DM0_generalist | 49.08 | |
| 10 | pi0 | 46.41 |
Interactive version: theaggregate.ai/benchmark?slug=robochallenge-table-30 · How It Works · Data refreshed daily, snapshot 2026-10-20.