Visual Question Answering on Mantis
81.57AccuracyMiMo-VL
Evaluation results
| Method | Links | |
|---|---|---|
| MiMo-VLParams=7B2025.11 | 81.57 | |
| MiMo-EmbodiedParams=7B2025.11 | 81.1 | |
| GPT-4oParams=–2025.11 | 75.6 | |
| Claude 3.7 SonnetParams=–2025.11 | 75.1 | |
| Qwen2.5-VLParams=7B2025.11 | 74.7 | |
| InternVL3Params=8B2025.11 | 72.8 |