MiMo-Embodied-7B: benchmark results
Provider: Xiaomi. Access: Open.
Aggregate Rating 1651 ± 31, rank #616 of 10916 rated models, from 36 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Causal-Plan-Bench - Affordance Postcondition | 41 | Accuracy (%) on the affordance postcondition (functional sta | 92.3 |
| RobotEQ-Video - Real Videos | 76.2 | Real-subset Avg (%; action-properness judgment: given an ego | 80.9 |
| EPIC-Bench - Embodied Compositional Attributes | 38.74 | Target localization score (0-100) on the embodied compositio | 73.9 |
| Causal-Plan-Bench - Spatial Precondition | 38 | Accuracy (%) on the spatial precondition (conditions that mu | 73.1 |
| Causal-Plan-Bench - Strategic Rationale | 28.4 | Open-ended strategic rationale (why the step serves the high | 69.2 |
| EPIC-Bench - Affordance Region | 8.71 | Manipulation score (0-100) on affordance region localization | 68.2 |
| EPIC-Bench - Visual Matching | 40.23 | Navigation score (0-100) on visual matching (localizing an o | 68.2 |
| RobotEQ-Video | 74.1 | A-Avg: mean of the synthetic-subset and real-subset averages | 64.9 |
| EPIC-Bench - Spatial-Related Attributes | 44.17 | Target localization score (0-100) on the spatial-attribute t | 64.8 |
| EMem-Bench - Dynamic Tracking | 35.6 | Success rate (%; the 1052 Dynamic Tracking episodes, acting | 62.5 |
| EPIC-Bench | 37.16 | Average score (0-100) over all samples, each sample scored b | 62.5 |
| Causal-Plan-Bench - Effects | 33.67 | Effects score: mean four-choice accuracy (%) after answer no | 61.5 |
Interactive version: theaggregate.ai/model?slug=mimo-embodied-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.