Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Prompt Harmfulness Classification on OAI
Loading...
87.4
F1 Score
HaloGuard 1.0
65.664
71.307
76.95
82.593
Jul 2, 2026
F1 Score
Updated 22d ago
Evaluation Results
Method
Method
Links
F1 Score
HaloGuard 1.0
Size=4B
2026.07
87.4
HaloGuard 1.0
Size=0.8B
2026.07
83.7
NemoGuard
Size=8B
2026.07
81
ShieldGemma
Size=27B
2026.07
80.5
PolyGuard-Qwen
Size=7B
2026.07
74.1
LlamaGuard4
Size=12B
2026.07
73.5
WildGuard
Size=7B
2026.07
72.1
Qwen3Guard-Gen
Size=8B
2026.07
68.8
Qwen3Guard-Gen
Size=4B
2026.07
68.3
Qwen3Guard-Gen
Size=0.6B
2026.07
66.5
Feedback
Search any
task
Search any
task