JSON Schema Bulk Sampler

Overview

Paste a JSON Schema and the bulk sampler produces N example records that conform to it - each with realistic values for strings, numbers, dates, and enums. Seed control gives reproducible output; an optional "vary fields" toggle makes each record more distinctive rather than near-identical.

It's for developers seeding test databases, generating fixture data for end-to-end tests, or producing demo payloads for API documentation. Reach for it when you need a hundred plausible-looking users, or when a stakeholder asks for "show me what 50 of these would look like" before agreeing to a schema.

How it works

The sampler walks the JSON Schema (Draft 2020-12 by default, with fallback support for Draft 7) and synthesises values based on type and constraints: string honours minLength, maxLength, pattern, and format (email, uuid, date-time); number respects minimum, maximum, multipleOf; enum and const produce only their allowed values.

Output is deterministic for a given seed - reusing the same seed produces the same N records. The "vary fields" mode adds locale-aware variation (different names, addresses, dates) without violating the schema constraints.

Examples

• A simple user schema:

  {
    "type": "object",
    "properties": {
      "id": {"type": "string", "format": "uuid"},
      "name": {"type": "string"},
      "age": {"type": "integer", "minimum": 18, "maximum": 99}
    },
    "required": ["id", "name"]
  }
  

• Sample output (N=2):

  [
    {"id": "9c5f...", "name": "Alex Morgan", "age": 34},
    {"id": "3a72...", "name": "Sam Reed",    "age": 52}
  ]
  

• Enum field:

  {"type": "string", "enum": ["pending","paid","refunded"]}
  -> randomly picks one per record
  

• Nested array of objects:

  {"type": "array", "items": {"$ref": "#/$defs/Item"}, "minItems": 3}
  

FAQ

Does it honour $ref?

Yes - internal $ref and $defs references within the same document resolve correctly. External $ref URLs aren't fetched.

Are negative number constraints supported?

Yes - exclusiveMinimum, exclusiveMaximum, and multipleOf are all respected during number synthesis.

What about format keywords that aren't built in?

Standard formats (email, uuid, date, date-time, uri, ipv4, ipv6) produce realistic values. Custom or extended formats fall back to a generic string that matches pattern if present.

Will every output validate against the schema?

Yes - the sampler only emits values that satisfy the schema's constraints. If you hit a case where output fails validation, that's a bug worth reporting.

Try JSON Schema Bulk Sampler