GUI Agent on AITZ

69.4SR

M2-Miner-7B

Updated 4d ago

Evaluation Results

Method
M2-Miner-7B 2026.02	69.4	81.3	-	-
M2-Miner-3B 2026.02	66.6	78.6	-	-
UI-TARS-7B* 2026.02	65.8	80.4	-	-
GUI-Owl-7B 2026.02	65.1	78.9	-	-
Learn from AITZ 2026.01	59.92	-	75.63	613
OS-Atlas-7B 2026.02	58.5	74.1	-	-
Agent-Dice 2026.01	57.1	-	74.72	237
Qwen2.5-VL-7B 2026.02	54.6	78.4	-	-
InfiGUI-R1 2026.02	52.9	70.7	-	-
Qwen2.5-VL-3B 2026.02	52.7	75.1	-	-
GUI-R1-7B 2026.02	50.5	56.8	-	-
Zero-Shot 2026.01	45.08	-	63.41	20
CL from all three 2026.01	42.05	-	65.78	4
Learn from GUI-Odyssey 2026.01	37.67	-	63.16	20
CL from AITZ and AndroidControl 2026.01	37.41	-	65.81	0
GPT-4o 2026.02	35.3	70	-	-
OdysseyAgent 2026.02	31.6	59.2	-	-
Learn from AndroidControl 2026.01	30.85	-	61.34	0
Aguvis-7B 2026.02	19	35.7	-	-
OS-Genesis-7B 2026.02	8.5	20	-	-