Synthetic Dataset Generator

Generate realistic synthetic datasets with correlated fields, built-in presets, custom schemas, deterministic seeding, and multiple output formats.

What does Synthetic Dataset Generator do?

This actor generates synthetic (fake) data on demand -- no web scraping required. It produces structured datasets using Faker.js for realistic data generation and Copycat for deterministic, reproducible output. Use it to create test data, seed databases, populate staging environments, or benchmark data pipelines.

Key features

• Built-in presets -- One-click generation for user profiles, companies, e-commerce products, and server log events
• Custom schemas -- Define your own field names and types to generate exactly the data shape you need
• Cross-field correlations -- Age tracks with salary, company size correlates with revenue, out-of-stock items have zero quantity
• Deterministic mode -- Set a random seed for identical output across runs (same seed + same config = same data every time)
• Multiple output formats -- JSON (Apify dataset), CSV, or NDJSON
• 16 locales -- Generate names, addresses, and phone numbers in English, German, French, Japanese, Chinese, and more
• Fast and cheap -- No proxy, no browser, no external API calls. Generates 10,000+ records per second on 256MB memory

Built-in presets

Preset	Fields	Example use case
User Profiles	id, name, email, phone, DOB, age, gender, address, job title, company, salary	Test user databases, CRM seed data
Companies	company ID, name, industry, employees, revenue, founded year, website, CEO	Business directory mock data
E-commerce Products	product ID, name, description, category, price, SKU, stock, rating, reviews	Product catalog testing
Log Events	timestamp, level, service, host, request ID, method, path, status code, response time	Log pipeline testing, SIEM demos

Custom schema

Define any combination of fields with the customSchema input (JSON array):

[
  { "name": "user_id", "type": "uuid" },
  { "name": "username", "type": "name" },
  { "name": "signup_date", "type": "datetime" },
  { "name": "plan", "type": "enum", "options": { "values": ["free", "pro", "enterprise"], "weights": [0.7, 0.2, 0.1] } },
  { "name": "monthly_spend", "type": "number", "options": { "min": 0, "max": 500 } }
]

Supported field types

string, integer, number, boolean, date, datetime, email, phone, address, name, first_name, last_name, company, url, uuid, city, state, zip, country, job_title, salary, sentence, paragraph, enum

For enum type, pass options.values (array of choices) and optionally options.weights (probability weights).

Input

Parameter	Type	Default	Description
preset	string	user_profiles	Built-in preset or custom for custom schema
recordCount	integer	100	Number of records to generate (1 to 500,000)
customSchema	string		JSON array of field definitions (only used when preset is custom)
locale	string	en	Language/region for generated data
seed	integer	0	Random seed for deterministic output (0 = random)
outputFormat	string	json	Output format: json, csv, or ndjson
enableCorrelations	boolean	true	Apply cross-field correlations for realistic data

Output

JSON format (default)

Records are pushed to the Apify dataset. Each record is a flat JSON object matching the selected preset or custom schema.

CSV / NDJSON format

The file is saved to the key-value store under the key OUTPUT. A summary record is also pushed to the dataset with download instructions.

Example output (user_profiles preset)

{
  "id": "a1abcd86-4c43-4300-864f-066b9f5e43eb",
  "first_name": "Lydia",
  "last_name": "MacGyver",
  "email": "Taryn8@gmail.com",
  "phone": "(444) 909-7300",
  "date_of_birth": "2005-01-20",
  "age": 21,
  "gender": "Female",
  "address": "8239 Johnston Shore",
  "city": "West Chanelleburgh",
  "state": "Utah",
  "zip": "00084",
  "country": "Holy See (Vatican City State)",
  "job_title": "Principal Applications Developer",
  "company": "Nienow-Gibson, Bruen and Mayer",
  "salary": 109866,
  "created_at": "2025-09-23T21:04:26.698Z"
}

Cost

This actor uses Pay Per Event pricing:

• $0.10 per actor start
• $0.0001 per data record generated

Example: Generating 10,000 user profiles costs $0.10 (start) + $1.00 (records) = $1.10 total.

Performance

Records	Approximate time	Memory
100	< 1 second	256 MB
1,000	~1 second	256 MB
10,000	~5 seconds	256 MB
100,000	~30 seconds	256 MB

Use cases

• Database seeding -- Populate development and staging databases with realistic test data
• API testing -- Generate request/response payloads for load testing and integration tests
• Data pipeline validation -- Feed synthetic data through ETL pipelines to verify transformations
• UI prototyping -- Fill dashboards and reports with realistic-looking data
• Machine learning -- Generate training data for models that need structured tabular input
• Demo environments -- Create convincing demo data without using real customer information

Need more features?

If you need additional field types, presets, or output formats, file an issue or get in touch. We actively maintain this actor and welcome feature requests.