Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
GUI Agent Interaction on MobileWorld GUI-Only
Loading...
55.6
SR
Gemini-3-Pro + UI-Ins-7B
5.784
18.717
31.65
44.583
Apr 13, 2026
SR
Updated 5d ago
Evaluation Results
Method
Method
Links
SR
Gemini-3-Pro + UI-Ins-7B
Framework Type=Agentic...
2026.04
55.6
GPT-5 + UI-Ins-7B
Framework Type=Agentic...
2026.04
54
Claude-4.5-Sonnet + UI-Ins-7B
Framework Type=Agentic...
2026.04
47.8
Doubao-1.5-UI-TARS
Framework Type=End-to-...
2026.04
26.3
MAI-UI-8B
Framework Type=End-to-...
2026.04
19.7
ClawGUI-2B
Framework Type=Ours
2026.04
17.1
UI-Venus-72B
Framework Type=End-to-...
2026.04
16.4
Qwen3-VL-235B-A22B
Framework Type=End-to-...
2026.04
12.8
Qwen3-VL-32B
Framework Type=End-to-...
2026.04
11.9
MAI-UI-2B
Framework Type=End-to-...
2026.04
11.1
Qwen3-VL-8B
Framework Type=End-to-...
2026.04
9.4
GUI-Owl-32B
Framework Type=End-to-...
2026.04
8.5
UI-Venus-7B
Framework Type=End-to-...
2026.04
8.5
GUI-Owl-7B
Framework Type=End-to-...
2026.04
7.7
Feedback
Search any
task
Search any
task