Text Query Safety on XSTest
95.01F1 ScoreWildGuard
Evaluation Results
| Method | Links | |
|---|---|---|
| WildGuard2026.06 | 95.01 | |
| YuFeng-XGuard-Reason-8B2026.06 | 94.61 | |
| SingGuard-4BParameters=4B2026.06 | 93.56 | |
| GPT-5.12026.06 | 91.69 | |
| GuardReasoner-VL-7B2026.06 | 91.47 | |
| SingGuard-2BParameters=2B2026.06 | 91.08 | |
| SingGuard-8BParameters=8B2026.06 | 90.61 | |
| Gemini3-Pro2026.06 | 89.62 | |
| Qwen3-VL-235B2026.06 | 89.5 | |
| Llama Guard 32026.06 | 88.52 | |
| Qwen3Guard-8BPolicy=loose2026.06 | 88.15 | |
| Qwen3Guard-8BPolicy=strict2026.06 | 87.96 | |
| ShieldGemma-9B2026.06 | 82.93 | |
| GraniteGuardian2026.06 | 72.69 |