CoreJSON

AI Mock Data Generator

This smart mock data generator uses a large language model to understand what field names mean (names, emails, addresses, amounts, dates and more) and produce realistic mock data — no hand-written mock rules required. It supports your own API key (BYOK): every request goes directly from your browser and never passes through this site.

Open in the workbench Runs locally · nothing is uploaded

How to use it

  1. 1

    Configure your API key

    Open "Model settings" and enter your OpenAI or compatible-platform API key. The key is stored only in your browser.

  2. 2

    Paste a skeleton or schema

    On the left, paste the JSON skeleton or schema you want to fill with mock values.

  3. 3

    Generate

    Click "Generate mock" and the AI fills the fields with data that matches their semantic meaning.

  4. 4

    Pass the data along

    Send the generated mock to the HTTP client for API testing without copy-pasting.

Key features

Semantic understanding

The AI reads the business meaning of each field and produces plausible data (real-looking names, cities and so on) rather than meaningless strings.

BYOK mode

Pure front-end connection to the model: sensitive business data never passes through this site, and there is no backend cost.

Wired into the toolchain

Fits the site workflow — format, then mock, then HTTP client — so data does not have to be copied around.

Works from a schema

Feed a JSON Schema or a plain skeleton and the model fills values in place, keeping your original field names and structure.

Output stays valid JSON

The generated result is parsed and formatted so the structure matches the skeleton you supplied.

OpenAI, Anthropic or compatible

Use the official OpenAI or Anthropic endpoints, or any OpenAI-compatible provider by supplying the full endpoint URL. The request is sent from the browser, so the endpoint must allow CORS.

Common errors & troubleshooting

  • ! For fields with a specific format, add a comment or a sample default in the skeleton; the AI uses that information when generating.
  • ! Bulk generation can hit the model token limit, so test with a small array first and then increase the count.
  • ! If the result does not match your intent, describe the expected values more concretely in the field names or comments.
  • ! The request is issued directly by the browser, so a third-party endpoint must support cross-origin (CORS) requests.

Frequently asked questions

Can I control how many mock records are generated?

Yes. Put an array with 2–3 elements in the input JSON, or hint at the quantity in a field description, and the AI expands or infers from that context.

Can my API key leak?

No. The API key lives only in your browser and is used to call the model vendor API directly; it is never sent to this site.

Can I generate mock data from a JSON Schema instead of a skeleton?

Yes. Paste either a JSON Schema or a plain JSON skeleton; the model fills the fields with semantically appropriate values while keeping your structure.

Is my business data uploaded anywhere?

Only to the model provider you configure, directly from your browser. This site has no backend and never receives your schema or the generated data.

Related tools

AI JSON Generator

Describe the data you want in plain language and let an LLM generate valid JSON. Bring your own API key (BYOK): requests go straight from your browser to the model provider.

JSON Schema to JSON

Generate example data from a JSON Schema. Supports const / enum / default / examples, local $ref (definitions / $defs), allOf / oneOf / anyOf and semantic format placeholders such as email, date-time and UUID.

HTTP Request Client

A lightweight, browser-based HTTP client. Configure method, URL, headers and a JSON body, send the request and inspect status, timing, headers and the formatted JSON response.

JSON Formatter

Format, beautify and validate JSON in your browser. Errors point to the exact line and column, and you can indent with 2 spaces, 4 spaces or a tab.