GUI Perception and Reasoning on MPR-GUI-Bench FR
86.9AUGemini-2.5-Pro
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Gemini-2.5-ProModel Category=Close-source LVLMs2025.11 | 86.9 | 64.5 | 81.6 | 92.9 | 81.4 | 65.3 | 88 | 76 | 79.6 | |
| Gemini-1.5-FlashModel Category=Close-source LVLMs2025.11 | 84.4 | 80.6 | 74 | 90.4 | 64.8 | 65 | 64 | 36 | 66 | |
| Keye-VL-7BModel Category=Open-source LVLMs2025.11 | 84.2 | 82.5 | 76.4 | 91 | 72.4 | 71.6 | 44 | 44 | 66.5 | |
| Qwen-2.5-VL-7B-InstructModel Category=Open-source LVLMs2025.11 | 81.7 | 83.6 | 80 | 91.3 | 76.5 | 79 | 72 | 72 | 80.3 | |
| Intern2.5VL-8BModel Category=Open-source LVLMs2025.11 | 77.1 | 83.9 | 75.6 | 88.5 | 72.7 | 76.5 | 80 | 52 | 73.5 | |
| MiniCPM-o 2.6Model Category=Open-source LVLMs2025.11 | 76.5 | 79 | 77.8 | 89.9 | 74.3 | 76.8 | 76 | 60 | 74.6 | |
| CogAgent-9BModel Category=GUI Agents2025.11 | 56.9 | 69.7 | 58.9 | 68 | 43.2 | 38.2 | 52 | 36 | 51 |