{
  "id": "llama-prompt-guard-2",
  "name": "Llama Prompt Guard 2",
  "category": "evals-observability",
  "summary": "Meta's small open-weight classifiers (86M and 22M) that flag prompt injection and jailbreak attempts in user input and untrusted content.",
  "description": "Binary benign/malicious labels; the 22M version (DeBERTa-xsmall) cuts latency and compute by about 75% versus 86M with minimal performance loss, per the model card. Used as LlamaFirewall's PromptGuard scanner. Gated weights under the Llama 4 Community License.",
  "url": "https://huggingface.co/meta-llama/Llama-Prompt-Guard-2-86M",
  "repo": "https://github.com/meta-llama/PurpleLlama",
  "docs": "https://github.com/meta-llama/PurpleLlama/blob/main/Llama-Prompt-Guard-2/86M/MODEL_CARD.md",
  "tags": [
    "prompt-injection",
    "jailbreak",
    "classifier",
    "open-weights",
    "meta"
  ],
  "license": "LicenseRef-Llama-4-Community-License",
  "pricing": "free",
  "status": "active",
  "related": [
    "llamafirewall",
    "llama-guard"
  ],
  "sources": [
    "https://github.com/meta-llama/PurpleLlama/blob/main/Llama-Prompt-Guard-2/86M/MODEL_CARD.md",
    "https://huggingface.co/meta-llama/Llama-Prompt-Guard-2-86M",
    "https://huggingface.co/meta-llama/Llama-Prompt-Guard-2-22M"
  ],
  "added": "2026-11-02",
  "updated": "2026-11-02",
  "last_verified": "2026-11-02",
  "submitted_by": "agentica-curator",
  "maintainer": "Meta",
  "longform": [],
  "links": {
    "html": "https://indexagentica.com/entries/llama-prompt-guard-2/",
    "markdown": "https://indexagentica.com/entries/llama-prompt-guard-2.md",
    "json": "https://indexagentica.com/api/entries/llama-prompt-guard-2.json",
    "source": "https://github.com/Drudley/indexagentica/blob/main/content/evals-observability/llama-prompt-guard-2.json"
  }
}