GUI Understanding on AndroidControl High
83.7Task Match Rate (TM)UI-TARS-7B
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| UI-TARS-7BModel Type=Open-source, Model Scale=7B2026.04 | 83.7 | 80.5 | 72.5 | |
| SeeClickModel Type=Open-source2026.04 | 82.9 | 62.9 | 59.1 | |
| UI-Copilot-7BModel Type=Open-source, Model Scale=7B2026.04 | 82.9 | 72.2 | 71.8 | |
| UI-AGILE-7BModel Type=Open-source, Model Scale=7B2026.04 | 80.1 | 61.9 | 60.6 | |
| UI-S1-7BModel Type=Open-source, Model Scale=7B2026.04 | 79.9 | 73.4 | 68.2 | |
| UI-AGILE-3BModel Type=Open-source, Model Scale=3B2026.04 | 78.6 | 60.7 | 56.8 | |
| AgentCPM-GUI-8BModel Type=Open-source, Model Scale=8B2026.04 | 77.7 | — | 69.2 | |
| NaviMaster-7BModel Type=Open-source, Model Scale=7B2026.04 | 72.9 | — | 54 | |
| UI-R1-7BModel Type=Open-source, Model Scale=7B2026.04 | 72.4 | 62.8 | 54.2 | |
| GUI-R1-7BModel Type=Open-source, Model Scale=7B2026.04 | 71.6 | 65.6 | 51.7 | |
| PAL-UI-7BModel Type=Open-source, Model Scale=7B2026.04 | 71.3 | 70.5 | 57.8 | |
| GPT-4oModel Type=Closed-source2026.04 | 66.3 | 0 | 20.8 | |
| OS-Genesis-7BModel Type=Open-source, Model Scale=7B2026.04 | 65.9 | — | 44.4 | |
| Aguvis-7BModel Type=Open-source, Model Scale=7B2026.04 | 65.6 | — | 54.2 | |
| Claude-CUModel Type=Closed-source2026.04 | 63.7 | 0 | 12.5 | |
| Qwen2.5VL-7BModel Type=Open-source, Model Scale=7B2026.04 | 62.2 | 72.5 | 52.7 | |
| PAL-UI-3BModel Type=Open-source, Model Scale=3B2026.04 | 60.4 | 58.7 | 49.3 | |
| GUI-R1-3BModel Type=Open-source, Model Scale=3B2026.04 | 58 | 56.2 | 46.6 | |
| UI-R1-3BModel Type=Open-source, Model Scale=3B2026.04 | 57.9 | 55.7 | 45.4 | |
| OS-Atlas-7BModel Type=Open-source, Model Scale=7B2026.04 | 57.4 | 54.9 | 29.8 | |
| OS-Atlas-4BModel Type=Open-source, Model Scale=4B2026.04 | 49 | 49.5 | 22.8 | |
| Qwen2.5VL-3BModel Type=Open-source, Model Scale=3B2026.04 | 47.8 | 46.5 | 38.9 |