domains / families / security / prompt-guard

prompt-guard

separate

Meta's 86M prompt injection and jailbreak detector. Filters malicious prompts and third-party data for LLM apps. 99%+ TPR, <1% FPR. Fast (<2ms GPU). Multilingual (8 languages). Deploy with HuggingFace or batch processing for RAG security.

When to use

• Need lightweight (86M params, <2ms latency)
• Filtering user inputs for jailbreaks
• Validating third-party data (APIs, RAG)
• Need multilingual support (8 languages)
• Budget constraints (CPU-deployable)
• **TPR**: 99.7% (in-distribution), 97.5% (OOD)
• **FPR**: 0.6% (in-distribution), 3.9% (OOD)
• **Languages**: English, French, German, Spanish, Portuguese, Italian, Hindi, Thai

Skill page

Source