MiMo-VL-7B-RL: benchmark results
Provider: Xiaomi. Released 2025-05-30. Access: Open.
Aggregate Rating 1586 ± 11, rank #1162 of 10916 rated models, from 240 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LogiScope-VQA - Personnel Safety Duty | 70 | Micro-averaged accuracy (%) | 100 |
| VirtueBench (1024 Frames) | 33.9 | Accuracy (%) over VirtueBench's 587 open-ended long-video qu | 92.3 |
| VISTA-Bench - Multimodal Perception | 70.7 | Accuracy (%) on the 300 multimodal perception questions (glo | 91.4 |
| OpenVLM Video - Video-MME (w/o subs) - Temporal Perception | 78.2 | Accuracy (%) | 88.9 |
| VISTA-Bench - Multimodal Reasoning | 59 | Accuracy (%) on the 300 multimodal reasoning questions (logi | 86.2 |
| SIS-Bench - Spatial Consistency | 72.3 | Accuracy (%; 195 spatial consistency questions; four-option | 86 |
| OpenVLM Video - MVBench - Action Count | 56.5 | Accuracy (%) | 84.7 |
| OpenVLM Video - MVBench - Moving Direction | 60 | Accuracy (%) | 83.3 |
| SCOPED - We-Math | 83.39 | Accuracy (%) | 83.3 |
| VISTA-Bench | 61.6 | Accuracy (%) over all 1,500 questions (300 perception, 300 r | 82.8 |
| VISTA-Bench - Unimodal Knowledge | 70.4 | Accuracy (%) on the 500 unimodal knowledge questions (natura | 82.8 |
| OpenVLM Video - Video-MME (w/o subs) - Action Recognition | 66.5 | Accuracy (%) | 81.9 |
Interactive version: theaggregate.ai/model?slug=mimo-vl-7b-rl · How It Works · Data refreshed daily, snapshot 2026-11-08.