Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Mobile GUI Agents on AndroidWorld 138 tasks (test)
Loading...
71.1
Success Rate
UI-Mem-8B
21.804
34.602
47.4
60.198
Feb 5, 2026
Success Rate
Updated 4d ago
Evaluation Results
Method
Method
Links
Success Rate
UI-Mem-8B
Params.=8B, Inference-...
2026.02
71.1
Seed1.8
Params.=-, Inference-t...
2026.02
70.7
Gemini-2.5-Pro
Params.=-, Inference-t...
2026.02
69.7
Step-GUI-8B
Params.=8B, Inference-...
2026.02
67.7
UI-Mem-8B
Params.=8B, Inference-...
2026.02
66.8
GUI-Owl-7B
Params.=7B, Inference-...
2026.02
66.4
UI-Tars-1.5
Params.=-, Inference-t...
2026.02
64.2
UI-Mem-4B
Params.=4B, Inference-...
2026.02
62.5
Seed1.5-VL
Params.=-, Inference-t...
2026.02
62.1
UI-Mem-4B
Params.=4B, Inference-...
2026.02
58.2
MAI-UI-2B
Params.=2B, Inference-...
2026.02
49.1
UI-Venus-7B
Params.=7B, Inference-...
2026.02
49.1
Qwen3-VL-8B
Params.=8B, Inference-...
2026.02
47.6
Qwen3-VL-4B
Params.=4B, Inference-...
2026.02
45.3
Qwen3-VL-2B
Params.=2B, Inference-...
2026.02
36.4
UI-Tars-1.5-7B
Params.=7B, Inference-...
2026.02
30
Ferret-UI Lite-3B
Params.=3B, Inference-...
2026.02
28
ScaleCUA-3B
Params.=3B, Inference-...
2026.02
23.7
Feedback
Search any
task
Search any
task