List your company

Prompt Attack Detection on Curated prompt-attack dataset

0.04LatencyPromptGuard
-0.19361.38322.964.5368Mar 26, 2026

Evaluation results

MethodLinks
PromptGuardModel Type=Encoder-based, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.030.0440.3215.7222.62
ProtectAIModel Type=Encoder-based, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.030.0432.1439.6235.49
gpt_oss_safeguardModel Type=Specialized LLM, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.030.5388.6473.5880.41
aws_prompt_attackModel Type=Propriertary, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.030.637.1437.1111.98
Qwen3Guard (0.6B)Model Type=Specialized LLM, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.03153.7554.0953.92
gemini-2.5-flash-liteModel Type=LLM Judge, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.031.4481.6581.1381.39
gemini-2.0-flash-lite-001Model Type=LLM Judge, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.031.5282.1486.7984.4
gemini-2.5-flashModel Type=LLM Judge, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.031.8577.389.9483.14
gemini-3-flash-previewModel Type=LLM Judge, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.032.0279.7891.8285.38
gpt-5-miniModel Type=LLM Judge, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.032.6789.883.0286.27
gpt-5.1Model Type=LLM Judge, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.034.0497.6678.6287.11
claude-haiku-4-5@20251001Model Type=LLM Judge, Decision threshold (τ)=0.5, Internal reasoning=Disabled2026.035.8883.5389.3186.32