Visual Counting on CountBench
90.2ScoreDatology
Evaluation Results
| Method | Links | |
|---|---|---|
| DatologyScale=4B2026.05 | 90.2 | |
| DatologyScale=2B2026.05 | 89.9 | |
| Qwen3.5-VLScale=4B2026.05 | 89.8 | |
| DatologyScale=1B2026.05 | 88.8 | |
| MAmmoTH-VLScale=2B2026.05 | 86.1 | |
| Qwen3.5-VLScale=2B2026.05 | 85.3 | |
| LoMoModel=Qwen3.5-9B, Evaluation Protocol=Standard Evaluation2026.05 | 85.01 | |
| LoMoModel=Qwen3.5-9B, Evaluation Protocol=Rendered Evaluation2026.05 | 83.98 | |
| MAmmoTH-VLScale=1B2026.05 | 83.5 | |
| MAmmoTH-VLScale=4B2026.05 | 82.9 | |
| BAGELSize=14B2026.04 | 82.5 | |
| Tuna-2Size=7B2026.04 | 81.7 | |
| Qwen3-VLScale=4B2026.05 | 81.3 | |
| Qwen3-VLScale=2B2026.05 | 80.4 | |
| InternVL3.5Scale=4B2026.05 | 80.2 | |
| Standard SFTModel=Qwen3.5-9B, Evaluation Protocol=Standard Evaluation2026.05 | 80.08 | |
| JanusFlowSize=1.3B2026.04 | 78.6 | |
| Tuna-RSize=7B2026.04 | 77.8 | |
| Ming-UniVisionSize=16B2026.04 | 76.8 | |
| LLaVA-OV2026.04 | 76.2 | |
| Qwen2.5-VLSize=7B2026.04 | 74.1 | |
| TunaSize=7B2026.04 | 73.5 | |
| InternVL3.5Scale=2B2026.05 | 71.7 | |
| Standard SFTModel=Qwen3.5-9B, Evaluation Protocol=Rendered Evaluation2026.05 | 71.66 | |
| HarmonSize=1.5B2026.04 | 67 | |
| InternVL3Scale=2B2026.05 | 66.2 | |
| Emu3Size=8B2026.04 | 65.2 | |
| TarSize=7B2026.04 | 64.2 | |
| Show-o2Size=7B2026.04 | 63.5 | |
| VILA-USize=7B2026.04 | 55.2 | |
| Janus-ProSize=7B2026.04 | 53.2 | |
| LoMoModel=LLaVA-OV1.5-8B, Evaluation Protocol=Standard Evaluation2026.05 | 51.12 | |
| Standard SFTModel=LLaVA-OV1.5-8B, Evaluation Protocol=Standard Evaluation2026.05 | 42.97 | |
| LoMoModel=LLaVA-OV1.5-8B, Evaluation Protocol=Rendered Evaluation2026.05 | 38.49 | |
| OneCatSize=9B2026.04 | 34.2 | |
| Standard SFTModel=LLaVA-OV1.5-8B, Evaluation Protocol=Rendered Evaluation2026.05 | 3.67 |