Mobile GUI Interaction on MobileWorld
55.6GUI-ONLY ScoreGemini-3-Pro + UI-Ins-7B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Gemini-3-Pro + UI-Ins-7BCategory=Agentic Framework2025.12 | 55.6 | 24.4 | 48.6 | 46.3 | |
| GPT-5 + UI-Ins-7BCategory=Agentic Framework2025.12 | 54 | 62.2 | 51.6 | 51.7 | |
| Claude-4.5-Sonnet + UI-Ins-7BCategory=Agentic Framework2025.12 | 47.8 | 37.8 | 50 | 43.8 | |
| MAI-UI-235B-A22BCategory=End-to-End Model2025.12 | 39.7 | 51.1 | 37.5 | 41.7 | |
| MAI-UI-32BCategory=End-to-End Model2025.12 | 36.2 | 46.7 | 30 | 37.3 | |
| MAI-UI-8BCategory=End-to-End Model2025.12 | 27.5 | 22.2 | 20 | 24.9 | |
| Doubao-1.5-UI-TARSCategory=End-to-End Model2025.12 | 26.3 | 32.4 | — | 20.9 | |
| UI-Venus-72BCategory=End-to-End Model2025.12 | 16.4 | 4.7 | — | 10.4 | |
| Qwen3-VL-235B-A22BCategory=End-to-End Model2025.12 | 12.8 | 4.4 | 5.4 | 9.5 | |
| Qwen3-VL-32BCategory=End-to-End Model2025.12 | 11.9 | 6.7 | 2.7 | 9 | |
| Qwen3-VL-8BCategory=End-to-End Model2025.12 | 9.4 | 0 | 0 | 5.5 | |
| GUI-Owl-32BCategory=End-to-End Model2025.12 | 8.5 | — | — | 5.5 | |
| UI-Venus-7BCategory=End-to-End Model2025.12 | 8.5 | 2.3 | — | 5.5 | |
| GUI-Owl-7BCategory=End-to-End Model2025.12 | 7.7 | — | — | 4.5 |