GUI Localization on GUIZoom-Bench hard-est
27.2Localization Accuracy (%)Qwen3-VL-32B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-VL-32BZoom Depth=d22025.12 | 27.2 | |
| Qwen3-VL-32BZoom Depth=d32025.12 | 25.5 | |
| Qwen3-VL-32BZoom Depth=d42025.12 | 25.5 | |
| Qwen3-VL-8BZoom Depth=d22025.12 | 19.9 | |
| Qwen3-VL-8BZoom Depth=d32025.12 | 18.8 | |
| Qwen3-VL-8BZoom Depth=d42025.12 | 18.2 | |
| UI-Venus-7BZoom Depth=d42025.12 | 16.2 | |
| UI-Venus-7BZoom Depth=d32025.12 | 16 | |
| UI-Venus-7BZoom Depth=d22025.12 | 14.6 | |
| UI-TARS-7BZoom Depth=d42025.12 | 12.9 | |
| UI-TARS-1.5-7BZoom Depth=d22025.12 | 12 | |
| UI-TARS-7BZoom Depth=d32025.12 | 11.8 | |
| UI-TARS-1.5-7BZoom Depth=d32025.12 | 11.2 | |
| UI-TARS-1.5-7BZoom Depth=d42025.12 | 10.4 | |
| Qwen3-VL-32BZoom Depth=d12025.12 | 10.1 | |
| Qwen2.5-VL-7BZoom Depth=d42025.12 | 9.5 | |
| Qwen2.5-VL-7BZoom Depth=d32025.12 | 9 | |
| UI-Venus-7BZoom Depth=d12025.12 | 8.7 | |
| UI-TARS-7BZoom Depth=d22025.12 | 8.4 | |
| Qwen3-VL-8BZoom Depth=d12025.12 | 8.1 | |
| Qwen2.5-VL-7BZoom Depth=d22025.12 | 8.1 | |
| Claude-4.5-SonnetZoom Depth=d12025.12 | 6.7 | |
| UI-TARS-1.5-7BZoom Depth=d12025.12 | 6.5 | |
| UI-TARS-7BZoom Depth=d12025.12 | 6.2 | |
| GPT-5Zoom Depth=d32025.12 | 4.8 | |
| GPT-5Zoom Depth=d42025.12 | 4.2 | |
| Qwen2.5-VL-7BZoom Depth=d12025.12 | 3.1 | |
| GPT-5Zoom Depth=d22025.12 | 2.8 | |
| GPT-5Zoom Depth=d12025.12 | 0.7 |