Provides the main interface for Guardrails. Can be used without configuration for string-based LLM apps and accepts a pydantic object for structured data use cases. Wraps LLM calls, orchestrates validation, and keeps track of call history.
02
Validation functionality
Runs Input/Output Guards in applications to detect, quantify, and mitigate specific types of risks. Validates the output against configured guardrails.
03
Structured data generation
Helps generate structured data from Language Models (LLMs). Uses either function calling for LLMs that support it or prompt optimization by adding the schema of the expected output to the prompt for LLMs that don't support function calling.
04
Guardrails Hub
Provides a collection of pre-built measures of specific types of risks called 'validators'. Allows users to find and contribute guardrails for LLM validation, mix and match guardrails, and build their own 'guard' that runs in the critical path in production.
05
Multiple validator support
Allows running multiple guardrails within a Guard. Users can install necessary guardrails from Guardrails Hub and create a Guard from the installed guardrails.
Your plan caps how many capabilities are shown — upgrade to see the full list
