List your company

Mobile GUI Agent Decision Making on AndroidWorld

82.05Success RateEAPO-8B
27.1941.432555.67569.9175Mar 20, 2025May 28, 2025Aug 5, 2025Oct 13, 2025Dec 21, 2025Feb 28, 2026May 9, 2026

Evaluation results

MethodLinks
EAPO-8BModel Category=Ours2026.0582.05———
EAPO-4BModel Category=Ours2026.0579.59———
EAPO-1.7B/2BModel Category=Ours2026.0576.36———
UI-TARS-2-2509Model Category=Closed-source Models2026.0573.38———
Claude-4-Sonnet-0929Model Category=Closed-source Models2026.0571.6———
DeepMiner-Mano-72BModel Category=Closed-source Models2026.0568.48———
Qwen3-VL-32BModel Category=Open-source Models2026.0563.7———
DeepMiner-Mano-7BModel Category=Closed-source Models2026.0562.3———
Qwen3-VL-235B-A22BModel Category=Open-source Models2026.0562———
Seed1.5-VL-250717Model Category=Closed-source Models2026.0561.75———
Claude-3.7-SonnetModel Category=Closed-source Models2026.0560.83———
V-DroidDecision=Verifier2025.0359.50.74—54.3
DART-GUI-7BModel Category=Open-source Models2026.0557.99———
OpenAI CUAModel Category=Closed-source Models2026.0557.63———
OpenAI CUA o3Model Category=Closed-source Models2026.0555.43———
GUI-Owl-7BModel Category=Open-source Models2026.0552.04———
OpenCUA-32BModel Category=Open-source Models2026.0551.66———
ARPOModel Category=Open-source Models2026.0549.31———
TianXi-Action-7BModel Category=Closed-source Models2026.0548.93———
ZeroGUIModel Category=Open-source Models2026.0547.52———
UI-TARS-72B-dpoModel Category=Open-source Models2026.0546.65———
OpenCUA-7BModel Category=Open-source Models2026.0545.1———
GPT-4oDecision=w/ CoT2025.0341.36.7—63.4
GPT-4oDecision=w/o CoT2025.0339.25.74—19.6
UI-TARS-7BDecision=w/ CoT2025.0333.210.6—75.9
UI-TARS-7BModel Category=Open-source Models2026.0533.04———
UI-TARS-7BDecision=w/o CoT2025.0329.30.97—14.6