Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multimodal Safety on MM-Safety
Loading...
99.32
F1 Score
LLaVAShield
24.388
43.8415
63.295
82.7485
Jun 22, 2026
F1 Score
Updated 1mo ago
Evaluation Results
Method
Method
Links
F1 Score
LLaVAShield
Category=Open-Source M...
2026.06
99.32
ShieldGemma-2
Category=Open-Source M...
2026.06
96.56
Qwen3-VL-235B
Category=Open-Source M...
2026.06
88.18
SingGuard-8B
Category=Our Models
2026.06
88.12
Gemini3-Pro
Category=Closed-Source...
2026.06
87.65
SingGuard-4B
Category=Our Models
2026.06
86.05
GPT-5.1
Category=Closed-Source...
2026.06
85.68
SingGuard-2B
Category=Our Models
2026.06
84.09
SafeGuard-VL
Category=Open-Source M...
2026.06
68.66
GuardReasoner-VL-7B
Category=Open-Source M...
2026.06
66.98
Qwen3-VL-8B
Category=Open-Source M...
2026.06
64.06
LlamaGuard4
Category=Open-Source M...
2026.06
61.47
Qwen3-VL-4B
Category=Open-Source M...
2026.06
61.03
LlamaGuard3-Vision-11B
Category=Open-Source M...
2026.06
51.06
LlavaGuard
Category=Open-Source M...
2026.06
27.27
Feedback
Search any
task
Search any
task