Safety Moderation on Harmbench
87.2F1 ScoreQwen3Guard-8B-Gen-strict
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3Guard-8B-Gen-strictStreaming=no, Moderation Policy=Gen-strict, Backbone Scale=8B2026.04 | 87.2 | |
| Qwen3Guard-4B-Gen-strictStreaming=no, Moderation Policy=Gen-strict, Backbone Scale=4B2026.04 | 86.7 | |
| BingoGuardParameter Count=8B2025.12 | 86.4 | |
| WildGuardParameter Count=7B2025.12 | 86.2 | |
| WildguardSize=7B2026.03 | 86 | |
| Roblox Guard 1.0Parameter Count=8B2025.12 | 85.7 | |
| Llama-Guard-3Size=8B2026.03 | 85 | |
| Qwen3Guard-Stream-8B-strictStreaming=yes, Moderation Policy=strict, Backbone Scale=8B2026.04 | 85 | |
| LlamaGuard3Parameter Count=8B2025.12 | 84.9 | |
| ShieldGemmaParameter Count=7B2025.12 | 84.8 | |
| Qwen3Guard-Stream-8B-looseStreaming=yes, Moderation Policy=loose, Backbone Scale=8B2026.04 | 84.7 | |
| Qwen3Guard-Stream-4B-strictStreaming=yes, Moderation Policy=strict, Backbone Scale=4B2026.04 | 84.3 | |
| Qwen3Guard-Stream-4B-looseStreaming=yes, Moderation Policy=loose, Backbone Scale=4B2026.04 | 83.6 | |
| GPT-4o2025.12 | 83.5 | |
| Llama3-StreamGuard-8BStreaming=yes, Backbone Scale=8B2026.04 | 83.3 | |
| Llama3-StreamGuard-3BStreaming=yes, Backbone Scale=3B2026.04 | 83.2 | |
| Qwen3Guard-Stream-0.6B-strictStreaming=yes, Moderation Policy=strict, Backbone Scale=0.6B2026.04 | 83.1 | |
| Llama3-StreamGuard-1BStreaming=yes, Backbone Scale=1B2026.04 | 82.4 | |
| MD-JudgeSize=7B2026.03 | 81 | |
| Qwen3Guard-Stream-0.6B-looseStreaming=yes, Moderation Policy=loose, Backbone Scale=0.6B2026.04 | 80.6 | |
| NemoGuardParameter Count=8B2025.12 | 78.6 | |
| FanarGuardSize=4B2026.03 | 77 | |
| PolyGuard-MinistralSize=8B2026.03 | 76 | |
| PolyGuard-QwenSize=7B2026.03 | 75 | |
| PolyGuard-SmolSize=0.5B2026.03 | 72 | |
| ShieldGemma-2bSize=2B2026.03 | 69 |