Visual Counting on CountBench 13 (test)
96.1ScoreGemini 3 Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini 3 ProModel Category=API call only2026.01 | 96.1 | |
| Molmo2-O-7BModel Category=Molmo2 family2026.01 | 95.1 | |
| Molmo2-4BModel Category=Molmo2 family2026.01 | 93.9 | |
| Molmo2-8BModel Category=Molmo2 family2026.01 | 93.7 | |
| PLM-8BModel Category=Open models2026.01 | 91.8 | |
| Molmo-72BModel Category=Molmo1 family2026.01 | 91.2 | |
| GPT-5Model Category=API call only2026.01 | 90.8 | |
| Gemini 2.5 ProModel Category=API call only2026.01 | 90.8 | |
| Qwen3-VL-8BModel Category=Open weights only2026.01 | 90.4 | |
| Eagle2.5-8BModel Category=Open weights only2026.01 | 90.2 | |
| Molmo-7B-OModel Category=Molmo1 family2026.01 | 89 | |
| Molmo-7B-DModel Category=Molmo1 family2026.01 | 88.5 | |
| GLM-4.1V-9BModel Category=Open weights only2026.01 | 88 | |
| Claude Sonnet 4.5Model Category=API call only2026.01 | 87.3 | |
| MolmoE-1BModel Category=Molmo1 family2026.01 | 87.2 | |
| GPT-5 miniModel Category=API call only2026.01 | 87.1 | |
| PLM-3BModel Category=Open models2026.01 | 87.1 | |
| Gemini 2.5 FlashModel Category=API call only2026.01 | 86.7 | |
| Qwen3-VL-4BModel Category=Open weights only2026.01 | 85.5 | |
| MiniCPM-V-4.5-8BModel Category=Open weights only2026.01 | 83.9 | |
| InternVL3.5-4BModel Category=Open weights only2026.01 | 82.2 | |
| Keye-VL-1.5-8BModel Category=Open weights only2026.01 | 81.6 | |
| InternVL3.5-8BModel Category=Open weights only2026.01 | 79.6 |