Share your thoughts, 1 month free Claude Pro on usSee more

Mobile Agent Interaction on MobileIAR English users

50.23Success Rate

GUI-owl-7B

Updated 13d ago

Evaluation Results

Method	Links
GUI-owl-7B 2025.08		50.23	84.2	44.63
Qwen3-VL-8B 2025.08		48.79	83.88	44.07
OS-Atlas-7B-Pro 2025.08		48.46	80.98	43.46
UI-TARS-7B-SFT 2025.08		44.48	72.57	40.69
UI-TARS-7B-SFT 2025.08		43.11	69.26	37.28
GUI-owl-7B 2025.08		42.99	82.71	38.13
UI-TARS-1.5-7B 2025.08		42.78	72.72	39.26
OS-Atlas-7B-Pro 2025.08		42.3	75.39	36.65
UI-TARS-7B-DPO 2025.08		41.58	70.91	37.73
UI-TARS-7B-DPO 2025.08		41.12	71.11	34.96
Qwen3-VL-8B 2025.08		40.75	77.34	36.64
GPT-4o + OCR model 2025.08		40.55	74.5	36.02
UI-TARS-1.5-7B 2025.08		40.19	72.06	34.02
GPT-4o + OCR model 2025.08		35.63	74.75	32.05
Qwen2.5-VL-7B 2025.08		30.57	40.67	27.7
Qwen-VL-max + OCR model 2025.08		20.37	81.82	17.56
Qwen-VL-max + OCR model 2025.08		19.86	79.91	16.69
Qwen2.5-VL-7B 2025.08		12.29	16.13	11.8
GLM-4v + OCR model 2025.08		3.21	73.14	2.47
GLM-4v + OCR model 2025.08		2.54	57.94	1.76