> ## Documentation Index
> Fetch the complete documentation index at: https://qwed-ai-mintlify-cfa51465.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# QWED vs Guardrails AI

> Compare QWED and Guardrails AI for LLM verification, AI agent security, schema validation, and deterministic output checking with side-by-side examples.

QWED and Guardrails solve different problems in an AI stack.

Guardrails helps you constrain structure. QWED helps you verify correctness.

## Core difference

| Tool       | Primary job                       | Best at                                          |
| ---------- | --------------------------------- | ------------------------------------------------ |
| Guardrails | Output structure and policy rules | JSON schemas, validators, formatting constraints |
| QWED       | Deterministic verification        | Math, logic, code, SQL, tool-call verification   |

## When Guardrails is enough

Use Guardrails when you need:

* Valid JSON or XML
* Required fields and schema checks
* Basic policy filters
* Controlled output formats for downstream parsing

## When you need QWED

Use QWED when you need:

* Formal verification for LLM outputs
* AI agent security before tool execution
* Verified tool calls and MCP security
* Deterministic checks for numbers, logic, code, and SQL

## Best-practice stack

Use both together:

1. Guardrails enforces the response format.
2. QWED verifies whether the contents are actually correct.

## Related docs

* [LLM verification with formal methods](/advanced/llm-verification)
* [QWED vs RAG for LLM verification](/advanced/qwed-vs-rag)
* [QWED vs Guardrails, RAG, and RLHF for LLM verification](/advanced/comparison)
* [QWED Open Responses: verified tool calls for AI agents](/open-responses/overview)