Visual Spatial Reasoning on BLINK Rel. Depth
94.4AccuracyMimo-Embodied
Evaluation results
| Method | Links | |
|---|---|---|
| Mimo-EmbodiedModel Category=Embodied VLMs2026.06 | 94.4 | |
| GPT-5.4Model Category=General-Purpose VLMs2026.06 | 87.9 | |
| Embodied-R1.5Model Category=Ours2026.06 | 87.1 | |
| Gemini-Robotics-ER-1.5Model Category=Embodied VLMs2026.06 | 85.9 | |
| Gemini-2.5-ProModel Category=General-Purpose VLMs2026.06 | 85.5 | |
| RoboBrain-2.0-7BModel Category=Embodied VLMs2026.06 | 79 | |
| Qwen3-VL-8BModel Category=General-Purpose VLMs2026.06 | 78.2 | |
| Pelican-VL-1.0Model Category=Embodied VLMs2026.06 | 78.2 | |
| Embodied-R1Model Category=Embodied VLMs2026.06 | 76.6 | |
| InternVL3.5-8BModel Category=General-Purpose VLMs2026.06 | 75 | |
| Molmo-D-7BModel Category=General-Purpose VLMs2026.06 | 75 | |
| VeBrainModel Category=Embodied VLMs2026.06 | 71.8 | |
| GPT-4oModel Category=General-Purpose VLMs2026.06 | 64.5 | |
| MagmaModel Category=Embodied VLMs2026.06 | 58.9 |