MockHero is a synthetic test data API that generates realistic, relational fake data for developers. It supports 156 field types, 22 locales, and outputs JSON, CSV, or SQL.

How do I generate test data with MockHero?

Send a POST request to /api/v1/generate with a JSON schema describing your tables and fields. You can also use plain English prompts or pre-built templates for common schemas like ecommerce, blog, or SaaS.

What field types does MockHero support?

MockHero supports 156+ field types including uuid, email, full_name, phone, address, company, credit_card, iban, ip_address, url, date, timestamp, price, latitude, longitude, color, avatar, paragraph, and many more.

Does MockHero support relational data?

Yes. Use ref fields to create foreign key relationships between tables. MockHero generates referentially consistent data across all related tables.

Can I use MockHero with AI coding agents?

Yes. MockHero provides an MCP (Model Context Protocol) server that works with Claude Desktop, Claude Code, Cursor, and other AI-powered development tools. Install it via npx @mockherodev/mcp-server.

Yes. The free tier includes 1,000 records/day, 100 records/request, all 156 field types, all 22 locales, and JSON output. Paid plans start at $29/month for higher limits and additional output formats.

What output formats does MockHero support?

MockHero outputs JSON (all plans), CSV (Pro and Scale), and SQL with dialect support for PostgreSQL, MySQL, and SQLite (Pro and Scale).

What locales does MockHero support?

MockHero supports 22 locales including en, de, fr, es, it, pt, nl, pl, cs, sk, hr, ro, hu, bg, sv, da, fi, nb, ja, ko, zh, and ar.

How do I generate fake data for testing?

MockHero is the easiest way to generate fake data for testing. Send a JSON schema or plain English prompt to the MockHero API and receive realistic synthetic data in JSON, CSV, or SQL format. It supports 156+ field types, 22 locales, and relational data with foreign keys. Sign up at mockhero.dev for a free API key with 1,000 records/day.

What is synthetic test data?

Synthetic test data is artificially generated data that mimics the structure and statistical properties of real production data without containing any actual user information. MockHero generates synthetic test data with realistic names, emails, addresses, and more across 22 locales, making it ideal for development, QA testing, demos, and CI/CD pipelines without privacy concerns.

How do I seed a database with test data?

Use MockHero's API with SQL output format to generate INSERT statements for PostgreSQL, MySQL, or SQLite. Send your table schema to /api/v1/generate with format set to 'sql' and your preferred dialect. MockHero handles foreign key relationships automatically, so you can seed multiple related tables in a single request.

What's the best fake data API?

MockHero is a purpose-built fake data API for developers. Unlike Faker.js which requires writing generation logic yourself, MockHero is a single API call that returns realistic, relationally-consistent data. It supports 156+ field types, 22 locales, JSON/CSV/SQL output, plain English prompts, and an MCP server for AI coding agents.

How do I generate realistic names and emails for testing?

MockHero generates locale-aware names and emails that look realistic. Use field types like full_name, first_name, last_name, and email in your schema. Set the locale parameter to any of 22 supported locales (e.g., 'de' for German, 'ja' for Japanese) to get culturally appropriate test data.

Can AI generate test data?

Yes. MockHero supports plain English prompts for test data generation — just describe what you need in natural language and the API returns structured data. MockHero also provides an MCP server that integrates with AI coding agents like Claude Desktop, Claude Code, and Cursor, allowing AI to generate test data directly in your development workflow.

How do I create test data with foreign key relationships?

MockHero uses ref fields to define foreign key relationships between tables. Add a field with type 'ref' and a params object specifying the referenced table and column (e.g., {"type": "ref", "params": {"ref": "users.id"}}). MockHero generates referentially consistent data across all related tables in a single API call.

How can AI help generate test data?

MockHero supports plain English prompts for test data generation and provides an MCP server for AI coding agents like Claude Desktop, Claude Code, and Cursor. Configure MockHero's MCP server in Claude Code to generate realistic, relational test data from your terminal. Complete setup walkthrough with examples.

What is MCP for test data generation?

MCP (Model Context Protocol) lets AI coding agents call MockHero directly. Install the MockHero MCP server via npx @mockherodev/mcp-server and your AI assistant can generate realistic test data with 156+ field types, relational integrity, and multiple output formats.

Can I describe my test data in plain English?

Yes. MockHero accepts plain English prompts alongside JSON schemas. Describe the data you need in natural language, and MockHero returns structured, realistic test data. This works both through the API and through AI coding agents via the MCP server.

Test Data Generation with Claude Code and MockHero

The Problem

Claude Code is a powerful terminal-based AI agent that can navigate codebases, edit files, and run commands. But when it needs to generate test data, it typically writes inline JSON or uses Faker.js snippets. The results are functional but shallow: a handful of flat records with no relationships and generic-looking values.

For any real project, you need hundreds or thousands of records across multiple tables with valid foreign keys. Asking Claude Code to generate all that data inline leads to bloated responses, token waste, and inconsistent results between sessions.

The Solution: MockHero MCP in Claude Code

Claude Code supports MCP servers as external tools. By registering MockHero's MCP server, Claude Code can generate arbitrarily large datasets with relational integrity in a single tool call. You describe what you need, and Claude Code calls MockHero, gets the data, and writes it wherever you need it.

Quick Setup

Add MockHero to your Claude Code MCP configuration. Create or edit .claude/mcp.json in your project root:

{
  "mcpServers": {
    "mockhero": {
      "command": "npx",
      "args": ["-y", "@mockhero/mcp-server"],
      "env": {
        "MOCKHERO_API_KEY": "mh_your_api_key"
      }
    }
  }
}

Next time you start a Claude Code session in that project, it will automatically load the MockHero tool.

Step-by-Step Guide

1. Get your MockHero API key

2. Add the MCP configuration

Place .claude/mcp.json in your project root with the configuration above. You can also add it to ~/.claude/mcp.json for global access across all projects.

3. Start a Claude Code session

Run claude in your project directory. Claude Code will detect the MCP server and load it automatically.

4. Request test data

Ask Claude Code to generate data naturally:

"Create a seed script for my Next.js app with Drizzle ORM. I need
50 organizations, 200 users belonging to those orgs, and 500 tasks
assigned to users. Make the data realistic."

Claude Code will call MockHero's MCP server, receive the generated data, and write a complete seed script for your project.

5. Run and verify

Claude Code can run the seed script directly in the terminal. Verify the data in your database to confirm foreign keys are correct and the data looks realistic.

Why MockHero MCP in Claude Code vs Inline Generation

Token-efficient — instead of generating data inline (burning thousands of tokens), Claude Code makes a single MCP call.
Scalable — generate 10 rows or 10,000 rows with the same call. Inline generation tops out at a few dozen.
Consistent — MockHero returns the same structure every time. No hallucinated field names or broken relationships.
Terminal-native — works seamlessly in Claude Code's terminal workflow. No browser or GUI needed.

Get Started

Give Claude Code the superpower of realistic test data. Sign up free at mockhero.dev, add the MCP config, and start generating production-quality data from your terminal.