GUI Agent Planning and Execution on AMEX (test)
77.3Success Rate (Gmail)MobileIPL
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| MobileIPLTraining Data=MobileIPL2025.05 | 77.3 | 71.8 | 80 | 68.4 | 85.3 | 71.3 | 73.5 | 82.1 | 71.8 | 74.29 | |
| Qwen2-VL-7BTraining Data=AMEX + CoaT2025.05 | 75.9 | 68.1 | 77.7 | 66.2 | 76.8 | 66.4 | 77.5 | 79.6 | 65.6 | 70.93 | |
| UGround-7BTraining Data=AMEX2025.05 | 70.9 | 68.8 | 72.7 | 63.7 | 77.7 | 67.7 | 63.7 | 80.1 | 67.6 | 69.12 | |
| UI-Tars-7BTraining Data=AMEX2025.05 | 67.7 | 70 | 71.8 | 63.8 | 71.5 | 67.7 | 77 | 86.4 | 72.8 | 70.33 | |
| AriaUI-MoETraining Data=AMEX2025.05 | 63.1 | 62.3 | 68.5 | 58.9 | 83 | 54.7 | 62.5 | 83.3 | 66.9 | 64.1 | |
| SphAgent-7BTraining Data=AITW + AMEX2025.05 | 62.4 | 68.1 | 76.3 | 71.9 | 68.6 | 67.3 | 77.6 | 66 | 64.1 | 69.14 | |
| SphAgent-7BTraining Data=AMEX2025.05 | 61.7 | 68.2 | 77.7 | 72 | 71.9 | 64.6 | 79.6 | 71.3 | 69.6 | 70.71 | |
| OS-Atlas-7BTraining Data=AMEX2025.05 | 61.1 | 73.5 | 77.9 | 61.6 | 75.2 | 66.4 | 71 | 75.9 | 72 | 70.33 | |
| Qwen2-VL-7BTraining Data=AMEX2025.05 | 58 | 70.1 | 76.6 | 63.8 | 79.4 | 66.8 | 67.8 | 80.2 | 76.6 | 69.01 | |
| MobileIPL-7BTraining Data=AMEX (Seed)2025.05 | 57 | 60.2 | 68.8 | 63.1 | 75 | 50.2 | 65.6 | 77.7 | 62.6 | 62.19 | |
| SphAgent-7BTraining Data=AITW2025.05 | 32.1 | 45.9 | 46.1 | 35.1 | 48.3 | 61.1 | 55.9 | 43.3 | 42.9 | 45.63 | |
| SeeClick-7BTraining Data=AITW+External2025.05 | 28.2 | 29.4 | 18.1 | 20 | 30 | 53.1 | 30.7 | 37.1 | 27.4 | 30.44 |