GUI Automation on GUI Odyssey
94.6Task Metric (TM)UI-TARS-7B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| UI-TARS-7B2026.04 | 94.6 | 90.1 | 87 | |
| AgentCPM-GUI-8B2026.04 | 90.8 | — | 75 | |
| UI-S1-7B2026.04 | 76.3 | 61.7 | 59.5 | |
| UI-Copilot-7B2026.04 | 74.5 | 63.8 | 57.2 | |
| Qwen2.5VL-7B2026.04 | 67.4 | 56.3 | 52.4 | |
| UI-R1-7B2026.04 | 67.1 | 41.3 | 43.5 | |
| Claude-CU (SoM)2026.04 | 60.9 | 0 | 3.1 | |
| OS-Atlas-7B2026.04 | 60.4 | 39.7 | 27 | |
| UI-Copilot-3B2026.04 | 52.4 | 36.8 | 35.8 | |
| UI-R1-3B2026.04 | 52.2 | 34.5 | 32.5 | |
| Qwen2.5VL-3B2026.04 | 37.4 | 26.5 | 26.7 | |
| ShowUI-2B2026.04 | 34.8 | 24.6 | 21.4 | |
| GPT-4o (SoM)2026.04 | 34.3 | 0 | 3.3 | |
| Qwen2VL-2B2026.04 | 24.4 | 12.2 | 12.6 | |
| OS-Genesis-7B2026.04 | 11.7 | — | 3.6 |