Visual Counting on PixmoCount
77.5AccuracyMiMo-Embodied
Evaluation Results
| Method | Links | |
|---|---|---|
| MiMo-EmbodiedParams=7B2025.11 | 77.5 | |
| CodeDance-7BModel Category=Open-Source MLLMs with Tools2025.12 | 77.1 | |
| MiMo-VLParams=7B2025.11 | 73.35 | |
| Qwen2.5-VL-72BModel Category=Open-Source MLLMs2025.12 | 62.3 | |
| InternVL3Params=8B2025.11 | 62 | |
| Llava-OneVision-72BModel Category=Open-Source MLLMs2025.12 | 60.7 | |
| Qwen2.5-VLParams=7B2025.11 | 60.7 | |
| Deepeyes-7BModel Category=Open-Source MLLMs with Tools, reproduced=true2025.12 | 57.2 | |
| Qwen2.5-VL-32BModel Category=Open-Source MLLMs2025.12 | 56 | |
| Llava-OneVision-7BModel Category=Open-Source MLLMs2025.12 | 54.4 | |
| GPT-4oParams=–2025.11 | 54.4 | |
| Claude 3.7 SonnetParams=–2025.11 | 53.5 | |
| Qwen2.5-VL-7BModel Category=Open-Source MLLMs2025.12 | 50.4 |