List your company

Visual Grounding on ScreenSpot V2 (Device/Input Modality Breakdown)

90.5Average AccuracyLPO-8B
53.68463.24272.882.358Jun 11, 2025

Evaluation results

MethodLinks
LPO-8BCategory=GUI-specific (RL)2025.0690.597.982.995.986.495.684.2
UI-TARS-72BCategory=GUI-specific (SFT)2025.0690.394.886.391.287.991.587.7
SE-GUI-7BCategory=GUI-specific (RL)2025.0690.3——————
Qwen2.5-VL-7BCategory=General open-source2025.0688.897.687.290.274.293.281.3
UGround-V1-7BCategory=GUI-specific (SFT)2025.0687.69583.39577.892.177.2
Aguvis-7BCategory=GUI-specific (SFT)2025.068695.577.395.477.99172.4
UI-TARS-2BCategory=GUI-specific (SFT)2025.0684.795.279.190.768.687.278.3
OS-Atlas-7BCategory=GUI-specific (SFT)2025.0684.195.275.890.763.690.677.3
Qwen2.5-VL-3BCategory=General open-source2025.0680.993.473.588.158.68871.4
GPT-4o + OmniParser-v2Category=Proprietary2025.0680.795.574.692.360.98859.6
OS-Atlas-4BCategory=GUI-specific (SFT)2025.0671.987.259.772.746.485.963.1
OperatorCategory=Proprietary2025.0670.547.341.590.280.392.884.3
Magma-8BCategory=GUI-specific (SFT)2025.0661.562.853.48057.967.547.3
SeeClick-9.6BCategory=GUI-specific (SFT)2025.0655.178.450.770.129.355.232.5