GUI Agent Automation on AndroidWorld (AW) (Online)
25.1Success RateVeriGUI-7B
Evaluation results
| Method | Links | |
|---|---|---|
| VeriGUI-7BCategory=Ours2026.04 | 25.1 | |
| UI-S1-7BCategory=Open-source 7B2026.04 | 22.7 | |
| Qwen2.5-VL-7BCategory=Open-source 7B2026.04 | 14.9 | |
| VeriGUI-3BCategory=Ours2026.04 | 12.6 | |
| UI-R1-3BCategory=Open-source 3B2026.04 | 10.9 | |
| Qwen2.5-VL-3BCategory=Open-source 3B2026.04 | 6.1 |