Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
UI Polishing on UIPolish Real
Loading...
85
VLM Score
GPT-5
0.76
22.63
44.5
66.37
Nov 11, 2025
VLM Score
Updated 27d ago
Evaluation Results
Method
Method
Links
VLM Score
GPT-5
Access=Closed-source VLM
2025.11
85
Claude-4-5-Sonnet-thinking
Access=Closed-source VLM
2025.11
81
UI2CodeN-9B-RL
Training=RL
2025.11
80
Claude-4-Sonnet-thinking
Access=Closed-source VLM
2025.11
78
UI2CodeN-9B-SFT
Training=SFT
2025.11
76
Claude-3.7-Sonnet-thinking
Access=Closed-source VLM
2025.11
75
Gemini-2.5-pro
Access=Closed-source VLM
2025.11
74
o4-mini
Access=Closed-source VLM
2025.11
65
Doubao-1.6-thinking-250715
Access=Closed-source VLM
2025.11
61
Doubao-1.5-thinking-vision
Access=Closed-source VLM
2025.11
51
Qwen3-VL-32B-Instruct
Access=Open-source VLM
2025.11
46
GLM-4.1V-9B-Thinking
Access=Open-source VLM
2025.11
42
Qwen3-VL-8B-Thinking
Access=Open-source VLM
2025.11
32.1
GPT-4o
Access=Closed-source VLM
2025.11
26
Qwen2.5-VL-72B
Access=Open-source VLM
2025.11
23
MiMo-VL-7B-SFT
Access=Open-source VLM
2025.11
17
Gemini-2.5-flash
Access=Closed-source VLM
2025.11
17
MiMo-VL-7B-RL
Access=Open-source VLM
2025.11
16
Kimi-VL-A3B-Instruct
Access=Open-source VLM
2025.11
14
Kimi-VL-A3B-Thinking
Access=Open-source VLM
2025.11
14
Qwen2.5-VL-7B
Access=Open-source VLM
2025.11
11
InternVL3-78B
Access=Open-source VLM
2025.11
10
InternVL3-9B
Access=Open-source VLM
2025.11
4
Feedback
Search any
task
Search any
task