Observability & evals

Meta's open safeguard models that moderate LLM prompts and responses, including Llama Guard 4.

Open-source Python library for adding programmable guardrails to LLM-based conversational applications

Guardrails, evaluation and pre-production red teaming for AI agents, RAG apps and chatbots.

Guardrails, red teaming and risk management platform for securing AI applications and agents.

AI security platform covering discovery, assessment, runtime protection and governance of AI apps and agents

Meta's open classifier models that detect prompt injection and jailbreak attacks on LLM inputs.

Discovers, tests, secures and governs every AI agent in an organization from one platform.

IBM model family that judges whether LLM inputs and outputs meet criteria such as jailbreak, harm and hallucination checks.

Microsoft AI service that detects harmful text and image content and screens user input for prompt attacks.