JSON to JSON Schema Generator

How to use this tool

Paste valid JSON, decide whether keys in the sample should be marked required, and choose Generate schema. Review the result, then copy it into your project.

Example

Input

{"id": 1, "name": "Ada", "tags": ["dev"]}

Output

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "type": "object",
  "properties": {
    "id": { "type": "integer" },
    ...
  },
  "required": ["id", "name", "tags"]
}

What does this tool do?

JSON Schema describes the shape of JSON data: which properties exist, what types they hold and which are required. This generator reads one sample document and infers a draft 2020-12 schema from it, including nested objects, arrays and primitive types. It is a starting point for structured-output definitions, API validation and documentation.

Common mistakes and limitations

A schema inferred from one sample cannot know which fields are optional, which strings are really dates or enums, or which numeric ranges are valid. Empty arrays produce no item type, null values only record null, and mixed array elements are merged into a combined type. Treat the output as a draft and refine it manually before relying on it.

What a schema adds beyond a sample

A sample shows one valid document; a schema states the rules every valid document must follow. With a schema, an application can reject a missing field before it causes a bug, and a language-model request can be asked to return data in a specific shape. The generated draft captures types and structure, which is the tedious part, and leaves the meaning to you.

{
  "type": "object",
  "properties": { "age": { "type": "integer" } },
  "required": ["age"]
}

Refine the draft by hand

After generating, add constraints a sample cannot reveal: enum lists for fixed choices, format for dates and emails, minimum and maximum for numbers, minLength for strings and additionalProperties false when unknown keys should be rejected. Check the required list against real data, since one sample marks every key it contains.

Related: JSON Formatter, LLM JSON Repair

From schema to code and config

If you need typed models rather than validation rules, generate Dart classes from the same sample. For configuration files, convert the JSON to YAML or back and compare the structure.

Related: JSON to Dart, YAML to JSON

Frequently asked questions