List your company

GUI Element Localization on ScreenSpot v2

92.8AccuracyKimi-VL-16B-MoE
15.11235.28155.4575.619May 29, 2025

Evaluation results

MethodLinks
Kimi-VL-16B-MoEBackbone=16B-MoE2025.0592.8
ViGoRL-7bBackbone=7b2025.0591
ViGoRL-3bBackbone=3b2025.0586.5
ViGoRL-3bBackbone=3b, Feedback Loop=Multi-turn RL2025.0586.1
UI-TARS-2BBackbone=2B2025.0584.7
Qwen2.5-VL-3B + Vanilla GRPOBackbone=3B, Training Protocol=Vanilla GRPO2025.0584.4
OS-Atlas-7BBackbone=7B2025.0584.1
Qwen2.5-VL-3B + SFT directBackbone=3B, Training Protocol=SFT direct2025.0580.6
Qwen2.5-VL-7BBackbone=7B2025.0573.6
OS-Atlas-4BBackbone=4B2025.0571.9
Qwen2.5-VL-3BBackbone=3B2025.0568.4
SeeClick2025.0555.1
GPT-4o2025.0518.1