GUI Automation on WorldGUI Augmented 1.0
83.5Success Rate (Office)Human*
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Human*2025.02 | 83.5 | 89.9 | 80.8 | 77.8 | 85.7 | 85.3 | |
| WorldGUI-AgentBackbone Model=GPT-5.1, Framework=WorldGUI-Agent2025.02 | 37.6 | 40.4 | 53.5 | 46.6 | 26.5 | 45.8 | |
| WorldGUI-AgentBackbone Model=Claude-3.5-Sonnet, Framework=WorldGUI-Agent2025.02 | 32.6 | 19.1 | 46.5 | 26.7 | 18.4 | 36 | |
| WorldGUI-AgentBackbone Model=GPT-4o, Framework=WorldGUI-Agent2025.02 | 24.3 | 11.2 | 35.4 | 15.6 | 12.2 | 26 | |
| CCUBackbone Model=Claude-3.5-Sonnet, Framework=CC solidarity2025.02 | 19.3 | 14.6 | 32.3 | 22.2 | 6.1 | 23.6 | |
| WorldGUI-AgentBackbone Model=Gemini-2.0, Framework=WorldGUI-Agent2025.02 | 17 | 9 | 29.3 | 11.1 | 10.2 | 19.1 | |
| Agent S22025.02 | 16.5 | 59.6 | 45.5 | 37.8 | 16.3 | 34.2 | |
| AssistGUIBackbone Model=GPT-4o, Framework=AssistGUI2025.02 | 16.1 | 7.9 | 20.2 | 11.1 | 8.2 | 16.5 | |
| UI-TARS-1.5Backbone Model=UI-TARS-1.52025.02 | 16.1 | 2.2 | 9.1 | 6.7 | 0 | 12.3 | |
| Plan-ActBackbone Model=GPT-4o, Framework=Plan-Act2025.02 | 10.1 | 2.3 | 11.1 | 2.2 | 2 | 8.5 | |
| Plan-ActBackbone Model=Gemini-2.0, Framework=Plan-Act2025.02 | 3.2 | 3.4 | 16.2 | 2.2 | 2 | 6.9 |