JSON Schema from Sample

Paste a JSON example to infer a JSON Schema Draft 2020-12. Prototype structured output, tool params, and API contracts fast.

Read the full guide: JSON Schema API Design Guide: From Sample Generation to LLM Validation →

Privacy: processed locally, never uploaded.

↓ Paste in the input area below to see results instantly

JSON sample

Paste sample JSON to infer a JSON Schema (Draft 2020-12).

Inferred schema

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "type": "object",
  "properties": {
    "name": {
      "type": "string"
    },
    "score": {
      "type": "number"
    },
    "tags": {
      "type": "array",
      "items": {
        "type": "string"
      }
    }
  },
  "required": [
    "name",
    "score",
    "tags"
  ],
  "additionalProperties": false
}

Notes

Inference rules

Types come from sample values; arrays use the first item; object keys are all required. Good for structured-output prototypes; review before production.

Paste a JSON example to infer a JSON Schema Draft 2020-12. Prototype structured output, tool params, and API contracts fast.

Quick start

Paste sample

Valid JSON object or array.
Copy schema

Review required fields and types for response_format, etc.

Limitations

Arrays use first item type only; all object keys marked required; unions need manual edits.

Typical Workflow

When designing API response structures, prepare a sample response first. Paste it into the tool, and the generated Schema can be directly used in OpenAPI documentation. For instance, an e-commerce product data structure with nested SKU arrays and complex attributes would take significant manual effort, while auto-generation saves 80% of the time.

For LLM applications, use this tool to create structured output templates rapidly. Have the AI generate ideal output samples first, then convert them to Schema for constraining future outputs. For example, extracting a consistent {location, temperature, unit} structure from natural language weather responses.

Examples

Object sample

Input

{"name":"Ada","score":98}

Output

JSON Schema with properties

FAQ

JSON5 support?

Strict JSON only.

Does the generated Schema require manual adjustments?

The basic structure is usually production-ready, but we recommend: 1) Checking if all fields are necessary 2) Adding description annotations 3) Adjusting constraints like number ranges. The tool highlights potential issues (e.g. anyOf types) but won't auto-remove redundant fields.