Grounding on RefCOCO (val)
92.8AccuracyCogVLM-G
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CogVLM-G2026.05 | 92.8 | — | |
| Qwen2.5-VLModel Size=72B2026.04 | 92.7 | — | |
| UNINEXT-HModel Scale=H2026.05 | 92.6 | — | |
| ONE-PEACE2026.05 | 92.6 | — | |
| InternVL3.5-8BParameters=8B2026.05 | 92.4 | — | |
| TextHawk22026.05 | 91.9 | — | |
| GR3D-8BParameters=8B2026.05 | 91.8 | — | |
| Qwen2-VL-7BParameters=7B2026.05 | 91.7 | — | |
| ForeSightModel Size=7B2026.04 | 91.42 | — | |
| Grounding-DINO-LModel Scale=L2026.05 | 90.6 | — | |
| Qwen2.5-VLModel Size=7B2026.04 | 90 | — | |
| Qwen2.5-VL-7BParameters=7B2026.05 | 90 | — | |
| DeepEyesModel Size=7B2026.04 | 89.8 | — | |
| InternVL3-2BParameters=2B2026.05 | 89.8 | — | |
| Qwen2.5-VL-3BParameters=3B2026.05 | 89.1 | — | |
| Qwen2.5-VLModel Size=7B, Reproduced=true2026.04 | 88.91 | — | |
| InternVL3.5-2BParameters=2B2026.05 | 88.7 | — | |
| AutoToolSize=7B2026.05 | 88.6 | — | |
| Shikra-7BParameters=7B2026.05 | 87 | — | |
| DeepEyesSize=7B2026.05 | 86.1 | — | |
| InternVL3-1BParameters=1B2026.05 | 85.8 | — | |
| InternVL3.5-1BParameters=1B2026.05 | 85.4 | — | |
| Qwen2.5-VL-7BSize=7B2026.05 | 83.4 | — | |
| LVRParam Size=7B2026.07 | — | 73.8 | |
| Qwen2.5-VLParam Size=3B2026.07 | — | 86.3 | |
| Qwen2.5-VLParam Size=7B2026.07 | — | 88.1 | |
| Qwen2.5-VLParam Size=32B2026.07 | — | 92.7 | |
| SaGe-3BParam Size=3B2026.07 | — | 89 | |
| SaGe-7BParam Size=7B2026.07 | — | 89.8 | |
| Vision-R1Param Size=7B2026.07 | — | 72.5 |