Gemini Robotics-ER 1.5: benchmark results
Provider: Google. Released 2025-09-25. Access: API.
Aggregate Rating 1716 ± 29, rank #282 of 10916 rated models, from 49 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Molmo2-ER Release - RoboSpatial-VQ | 79.3 | Benchmark score (%) | 100 |
| HELM Robo-Reward-Bench - Berkeley Rpt Converted Externally To Rlds | 0.44 | Absolute Error | 91.7 |
| Molmo2-ER Release - Overall Avg. | 61.3 | Benchmark score (%) | 90.9 |
| HELM Robo-Reward-Bench - Iamlab Cmu Pickup Insert Converted Externally To Rlds | 0.51 | Absolute Error | 87.5 |
| HELM Robo-Reward-Bench - Tokyo U Lsmo Converted Externally To Rlds | 0.52 | Absolute Error | 87.5 |
| Molmo2-ER Release - Point-Bench | 73.3 | Benchmark score (%) | 81.8 |
| Molmo2-ER Release - Where2Place | 59 | Benchmark score (%) | 81.8 |
| EMem-Bench - Dynamic Tracking | 47.6 | Success rate (%; the 1052 Dynamic Tracking episodes, acting | 81.2 |
| Molmo2-ER Release - OpenEQA | 55 | Benchmark score (%) | 77.3 |
| HELM Robo-Reward-Bench - Droid | 0.76 | Absolute Error | 77.1 |
| EMem-Bench | 52.4 | Success rate (%; unweighted mean of the four task-family suc | 75 |
| EMem-Bench - Passive Observation | 60.2 | Success rate (%; the 1036 Passive Observation episodes, find | 75 |
Interactive version: theaggregate.ai/model?slug=gemini-robotics-er-1-5 · How It Works · Data refreshed daily, snapshot 2026-11-08.