LLM Guard
Visit pageComprehensive input/output scanners for LLM apps. MIT license.
Open-Source and Frameworks
LLM Guard
Open-source security toolkit for LLM-powered applications.
About LLM Guard
LLM Guard provides input scanners (anonymization, prompt injection, ban substrings) and output scanners (sensitive content, bias, factual checking). Drop-in for Python LLM applications. Maintained by Protect AI.
Notable open-source projects and reference frameworks used by enterprises and consultancies to harden AI deployments.
Products
LLM Guard products and platform components
Direct links to the vendor's product pages. Last reviewed 2026-05-07.
CWS engagement
How CWS works with LLM Guard
CWS helps customers evaluate, deploy, and operate LLM Guard products as part of an AI security program. Engagements span vendor selection, proof-of-concept design, integration with existing controls, day-2 operations, and exit planning if the fit changes over time.
CWS does not resell LLM Guard. The recommendation is honest, evidence-based, and tied to the customer's posture gaps — not to channel economics.
Engage CWS on LLM Guard Related — same category
More vendors in Open-Source and Frameworks
NVIDIA NeMo Guardrails
Open-source toolkit for adding programmable guardrails to LLM apps.
View profileGarak
Open-source LLM vulnerability scanner.
View profilePromptfoo
Open-source LLM evaluation, red teaming, and security testing.
View profilePyRIT
Microsoft's open-source Python Risk Identification Toolkit for GenAI.
View profileGiskard
Open-source LLM testing framework with hosted hub.
View profileNot sure if LLM Guard fits your gaps?
The free AI Posture Check scores your security across six dimensions in 10 minutes. Use the result to shortlist vendors that fit your actual posture — not the loudest demo.
Take the AI Posture Check