Visual Spatial Instruction on VSIBench
56.1AccuracyEmbodied-R1.5
Evaluation Results
| Method | Links | |
|---|---|---|
| Embodied-R1.5Model Category=Ours2026.06 | 56.1 | |
| InternVL3.5-8BModel Category=General-Purpose VLMs2026.06 | 55.7 | |
| Qwen3-VL-8BModel Category=General-Purpose VLMs2026.06 | 55.1 | |
| Pelican-VL-1.0Model Category=Embodied VLMs2026.06 | 52.8 | |
| GPT-5.4Model Category=General-Purpose VLMs2026.06 | 52.6 | |
| Mimo-EmbodiedModel Category=Embodied VLMs2026.06 | 48.5 | |
| GPT-4oModel Category=General-Purpose VLMs2026.06 | 43.6 | |
| Gemini-2.5-ProModel Category=General-Purpose VLMs2026.06 | 43.4 | |
| Gemini-Robotics-ER-1.5Model Category=Embodied VLMs2026.06 | 39.9 | |
| VeBrainModel Category=Embodied VLMs2026.06 | 36.2 | |
| RoboBrain-2.0-7BModel Category=Embodied VLMs2026.06 | 36.1 | |
| Embodied-R1Model Category=Embodied VLMs2026.06 | 26.6 | |
| MagmaModel Category=Embodied VLMs2026.06 | 17.1 | |
| Molmo-D-7BModel Category=General-Purpose VLMs2026.06 | 6.7 |