How do I validate JSON against a schema?

Paste your JSON data in the left editor and your JSON Schema in the right editor, then click Validate. The tool checks every constraint in the schema and reports all validation errors with their exact paths.

What schema features are supported?

The validator supports all JSON Schema Draft 2020-12 features including type checking, required fields, pattern matching, minimum/maximum constraints, array item validation, allOf/anyOf/oneOf combinators, and $ref references.

What do the validation errors mean?

Each error shows the data path (where the error is), the schema path (which rule failed), and a human-readable message explaining what went wrong. For example: /users/0/email must match format "email".

Can I validate API request bodies with this tool?

Yes. If you have an OpenAPI/Swagger spec, extract the relevant schema definition, paste it in the Schema editor, then paste a sample request body in the JSON editor to verify it meets all requirements.

Can I validate LLM structured output with this tool?

Yes. Paste the model output on one side and the schema you expect on the other. This works well for OpenAI Structured Outputs, Anthropic tool inputs, MCP tool schemas, and any agent workflow where downstream code depends on a strict JSON shape.

Validador JSON Schema

Valida documentos JSON contra um JSON Schema

JSON Data

Loading editor...

JSON Schema

Loading editor...

Enforce a JSON contract, not just JSON syntax

Syntax validation tells you the document parses. Schema validation tells you it actually has the fields, types, and constraints your code expects. That matters most when JSON is crossing a boundary — between services, between a model and your code, or between a partner system and yours — where a missing field becomes a production incident.

Use the validator when you need to

Validate API request and response bodies

Check that incoming and outgoing payloads match the contract before they reach business logic.

Validate AI structured outputs

Confirm that model responses, tool calls, or agent JSON match the shape your downstream code requires.

Run schema checks in CI on fixtures

Fail builds when sample payloads or fixtures drift away from the documented contract.

How to validate JSON against a schema

1
Paste your JSON document into the data editor.
2
Paste your JSON Schema (Draft 2020-12) into the schema editor.
3
Click Validate and review the path-by-path errors if validation fails.

Keep going

Generate a schema from a sample

Skip writing a schema from scratch — generate one from a known-good document.

Validate LLM output

Combine repair and schema validation when the input is model JSON.

Just check syntax

Confirm the document parses before running schema validation against it.

Build a provider schema

Wrap a schema for OpenAI, Anthropic, or MCP integrations.

Common schema-validation workflows

Lock down an API contract

Reject any request body that doesn't match the schema before it touches downstream services.

Verify a model output

Validate structured AI responses against the schema your handler expects.

Catch fixture drift in CI

Run validation on test fixtures so changes to the contract surface immediately.

Ferramentas relacionadas

Structured Output Schema Builder

Generate OpenAI, Anthropic, MCP, or plain JSON Schema wrappers from sample output

Gerador de JSON Schema

Gera JSON Schema a partir de dados JSON de exemplo

LLM JSON Output Validator

Repair and validate model JSON output against a schema

MCP Tool Schema Validator

Validate MCP tool definitions and verify `inputSchema` before deployment

Perguntas frequentes

Cole seus dados JSON no editor da esquerda e o JSON Schema no editor da direita, depois clique em Validate. A ferramenta verifica todas as restrições do schema e lista os erros de validação com os caminhos exatos.

O validador suporta todos os recursos do JSON Schema Draft 2020-12, incluindo checagem de tipos, campos required, validação por pattern, restrições minimum/maximum, validação de itens de array, combinadores allOf/anyOf/oneOf e referências $ref.

Cada erro mostra o data path (onde está o problema nos dados), o schema path (qual regra falhou) e uma mensagem legível explicando o que deu errado. Exemplo: /users/0/email must match format "email".

Sim. Se você tiver uma spec OpenAPI/Swagger, extraia a definição de schema relevante, cole no editor de Schema e depois cole um body de exemplo no editor de JSON para verificar se ele atende a todos os requisitos.