Loading GraphQL Schema Validator...
Please wait a moment

How to Validate GraphQL Schemas - Step by Step Guide

Step 1: Input Your GraphQL Schema

Paste your GraphQL schema definition, upload a .graphql or .gql file, or use the sample data to test the validator.

  • •Paste schema directly from your code editor
  • •Upload .graphql or .gql files
  • •Click "Sample" to load example schema

Example GraphQL Schema

type User {
  id: ID!
  name: String!
  email: String!
  posts: [Post!]!
}

type Post {
  id: ID!
  title: String!
  content: String!
  author: User!
}

Step 2: Automatic Schema Validation

About half a second after you stop typing, the schema is built with graphql-js and checked against the rules of the GraphQL specification.

  • •Validates GraphQL schema syntax
  • •Verifies type definitions and references
  • •Reports the line and column of the first problem
  • •Checks interface and union types and default values

Step 3: Review Validation Results

View validation results with detailed error messages and schema statistics.

  • •Success indicator for valid schemas
  • •Error messages with line numbers
  • •Schema statistics and type counts

Step 4: Fix and Re-validate

Fix the problem in the editor and the check runs again. Copy and Download give you your schema exactly as typed.

  • •Re-checked as you edit
  • •Download as schema.graphql
  • •Copy to clipboard

Frequently Asked Questions

How do I validate a GraphQL schema?

Paste your type definitions into the editor or upload a .graphql or .gql file. About half a second after you stop typing, the schema is built with graphql-js, the reference GraphQL implementation. You either get a green "Valid Schema" with counts of your types and fields, or the first problem with its line, column and the line itself marked with a caret.

What does the validator actually check?

Everything graphql-js checks when it builds a schema: syntax, types that are used but never defined, duplicate type, field, argument and enum value names, unknown directives, interfaces that a type claims to implement but doesn't fully provide, unions that include non-object types, input types used as output (and the other way round), a missing Query type, and default values that don't fit their type, such as an Int default past 2147483647. It does not run resolvers, lint naming style, or compare against a previous version of your schema.

Why does it say Query root type must be provided?

The GraphQL spec requires every schema to have a query root, so a file that only defines a few types fails on its own. If you are checking one module of a bigger schema, add a placeholder such as type Query { _empty: String } while you check it.

Does it work with Apollo Federation or AWS AppSync schemas?

Only partly. Plain SDL from Apollo Server, GraphQL Yoga and similar servers checks fine. Federation subgraphs use directives like @key and @external, and AppSync adds scalars like AWSDateTime and directives like @aws_auth. Standard GraphQL doesn't know those, so the validator reports them as unknown until you declare them in the file (for example directive @key(fields: String!) repeatable on OBJECT | INTERFACE, or scalar AWSDateTime).

I pasted a query and got an error. Why?

This page checks schema definitions, not operations. A query, mutation or fragment has no type definitions to validate, so the page tells you so instead of guessing. To tidy up a query, use the GraphQL Formatter or GraphQL Minifier.

Is my schema uploaded anywhere?

No. The checking runs in your browser with the graphql-js library, and nothing you paste or upload is sent to our server. Copy and Download give you back exactly the text in the editor.

Working on queries rather than schemas? Try the GraphQL Formatter or the GraphQL Minifier.