Guardrails AI

Validation and guardrails library for LLM applications.

Visit site GitHub
Open source Self-hostable free
GuardrailsPlatform engineer
LicenseApache-2.0
PriceFree
UsageNo model cost
Model supportmulti
Languagespython
GitHub stars7.1k
Last activity2026-06-30
Verified2026-06-28
Verified byseed

A library for adding structural validation, output formatting, and safety guardrails to LLM applications, with a registry of reusable validators.

Strengths

  • Rich validator registry for structure, safety, and correctness
  • Structural validation ensures outputs match expected schemas

Tradeoffs

  • Overhead in defining validators for custom use cases
  • Not a complete security solution on its own