Navigation on RealMobile (test)
85Success RateGemini 3.1 Pro
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Gemini 3.1 Pro2026.06 | 85 | 89.6 | |
| Seed 2.0 Pro2026.06 | 80 | 88.1 | |
| Xiaomi-GUI-0-30B-A3Bparameters=30B-A3B2026.06 | 72 | 85.8 | |
| Seed 1.82026.06 | 65 | 82.4 | |
| Claude Opus 4.72026.06 | 60 | 74.8 | |
| Gemini 3.1 Flash2026.06 | 58 | 72.4 | |
| Claude Opus 4.62026.06 | 33 | 56.7 | |
| MAI-UI-8Bparameters=8B2026.06 | 33 | 50.8 | |
| GUI-Owl-1.5-32B-Thinkingparameters=32B, mode=Thinking2026.06 | 31 | 51.7 | |
| GUI-Owl-1.5-8B-Thinkingparameters=8B, mode=Thinking2026.06 | 26 | 39 | |
| GUI-Owl-1.5-8B-Instructparameters=8B, mode=Instruct2026.06 | 25 | 44 | |
| UI-TARS-1.52026.06 | 24 | 40.5 | |
| GUI-Owl-1.5-32B-Instructparameters=32B, mode=Instruct2026.06 | 22 | 40.6 | |
| UI-Venus-1.5-30B-A3Bparameters=30B-A3B2026.06 | 21 | 44.6 | |
| UI-Venus-1.5-8Bparameters=8B2026.06 | 16 | 41.6 | |
| Step-GUI-8Bparameters=8B2026.06 | 15 | 32.8 |