GUI Grounding on Screenspot Desktop
95.9Text AccSE-GA
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SE-GAModel Size=7B2026.05 | 95.9 | 76.4 | |
| Qwen2.5VLModel Size=72B2026.05 | 95.3 | 74.3 | |
| AguvisModel Size=7B2026.05 | 93.8 | 67.1 | |
| Aria-UIModel Size=23B2026.05 | 93.3 | 64.3 | |
| OS-AtlasModel Size=7B2026.05 | 91.8 | 62.9 | |
| OmniParserBackbone=GPT-4V, Zero-shot=true2024.11 | 91.3 | 63.6 | |
| UI-TARSModel Size=72B2026.05 | 89.7 | 88.6 | |
| Qwen2.5VLModel Size=7B2026.05 | 85.4 | 67.8 | |
| UGroundSize=7B, #Train=1.3M, Zero-shot=true2024.11 | 82.5 | 63.6 | |
| UGroundModel Size=7B2026.05 | 82.5 | 63.6 | |
| ShowUI-GSize=2B, #Train=119K, Zero-shot=true, Training Mode=Grounding data only2024.11 | 81.8 | 59 | |
| ShowUISize=2B, #Train=256K, Zero-shot=true2024.11 | 76.3 | 61.1 | |
| CogAgentSize=18B, #Train=400K, Zero-shot=true2024.11 | 74.2 | 20 | |
| SeeClickSize=9.6B, #Train=364K, Zero-shot=true2024.11 | 72.2 | 30 | |
| SeeClickModel Size=9.6B2026.05 | 72.2 | 30 | |
| FuyuSize=8B, Zero-shot=true2024.11 | 33 | 3.6 | |
| GPT-4oModel Size=-2026.05 | 21.1 | 23.6 | |
| Qwen2-VL-2BSize=2B, Zero-shot=true2024.11 | 1.4 | 9.3 |