List your company

Mobile UI Reasoning on AMEX

76.2Gmail AccuracyCoME
26.2839.2452.265.16Feb 27, 2026

Evaluation results

MethodLinks
CoME#Params=5B2026.0276.272.68164.381.263.272.678.466.972.61
OS-Atlas#Params=7B2026.0274.469.774.66480.763.665.383.362.369.99
UGround#Params=7B2026.0270.968.872.763.777.767.773.780.167.669.12
UITars#Params=7B2026.0267.77071.863.871.567.77786.472.870.33
Qwen2VL-MoE#Params=3B, hybrid-capabilities reasoning finetuned=true2026.0265.163.757.8637860.370.276.561.164.56
AriaUI#Params=3.9B2026.0263.162.368.558.98354.762.583.366.964.1
SphAgent#Params=7B, reported_from=Chai et al., 20242026.0261.768.277.77271.964.679.671.369.670.71
UITars#Params=2B2026.0259.447.955.45365.162.364.761.758.154.43
Qwen2VL#Params=7B, hybrid-capabilities reasoning finetuned=true2026.0257.658.456.547.364.266.360.972.854.857.99
ShowUI#Params=2B2026.0252.233.668.855.351.857.451.969.750.247.74
UI-S1#Params=7B2026.0247.362.26654.138.175.372.780.263.660.57
DeepSeekVL2#Params=4.5B, hybrid-capabilities reasoning finetuned=true2026.0243.136.952.542.242.751.14761.738.942.22
MolmoE#Params=1B, hybrid-capabilities reasoning finetuned=true2026.0238.728.628.127.245.322.837.638.23431.56
Qwen2VL#Params=2B, hybrid-capabilities reasoning finetuned=true2026.0237.435.240.133.75047.541.856.740.538.53
GUI-R1#Params=3B2026.0237.150.658.247.121.654.754.656.154.448.53
GUI-R1#Params=3B2026.0236.85364.649.420.668.664.568.562.352.47
SeeClick#Params=7B, reported_from=Chai et al., 20242026.0228.229.418.1203053.130.737.127.430.44