Metasyn

Metasyn is a self-hosted Python package for generating synthetic tabular data from dataframes. It is best suited to researchers and public organizations working with sensitive datasets who want a reproducible workflow they can extend. Its strengths are dataframe support, auditable model settings and flexible exports; synthetic output still requires careful review.

Overview

Users fit a MetaFrame to a dataframe, save the model and settings as JSON, then generate new rows. That saved configuration supports an auditable, repeatable workflow rather than a one-off data export. Metasyn handles missing values, categorical fields, unique keys and structured strings such as postcodes or identifiers, making it relevant to varied tabular datasets.

By default, its model excludes multivariate information to reduce risks including identity, attribute and group disclosure. A disclosure-control plugin is also offered. Neither approach makes generated data automatically safe: output needs human review and is not guaranteed to be free of sensitive information. Metasyn is therefore a tool for responsible data preparation, not a substitute for privacy assessment.

Key features

Support for both Polars and pandas makes the workflow usable with two common Python dataframe libraries. Faker supplies localized names, email addresses, phone numbers and other data, while a plugin system lets users and organizations add extensions. That flexibility suits teams with specialized needs, though it makes Metasyn a better fit for Python-capable users than for people seeking a standalone point-and-click application.

Generated dataframes can be exported to CSV, Parquet, Excel and other formats. The package is available through pip from PyPI or directly from GitHub for the development version; versioned Docker containers and a command-line interface provide additional ways to build programs with Metasyn. API access is supported, and deployment is self-hosted. Data masking and data subsetting are not supported.

Pricing

Metasyn is free: its sole Metasyn plan costs nothing and is an open-source Python package under the MIT License. The license permits free use, modification, distribution and sublicensing subject to its conditions. There is no free trial because the package is already free. No paid tier or seat-based plan is described.

The free package gives users the software without a subscription, but they take responsibility for running it themselves and reviewing generated data. Bug reports and technical questions can go through GitHub issues, and users can contact Raoul Schram.

Platforms

Metasyn supports Linux, macOS and Windows, as well as self-hosted deployment. Its Python-package workflow and command-line option favor users comfortable building or running software in their own environment.

Who it's for

Public organizations, research groups and individual researchers who own sensitive datasets are the intended audience. It is a strong fit when teams need reproducible synthetic tabular data, dataframe-based work and control over deployment. It is less suitable for users who need masking or subsetting, or who expect synthetic output to remove the need for human privacy review.

Pros and cons

• Reproducible workflow: model and settings can be saved as auditable JSON, helping teams repeat and inspect their generation process.
• Broad tabular coverage: missing values, categories, unique keys and structured strings are supported, alongside Polars and pandas.
• Flexible output and extension: CSV, Parquet, Excel and other exports, plus plugins, accommodate varied workflows.
• Privacy remains a human responsibility: default exclusion of multivariate information and a disclosure-control plugin reduce some risks, but do not guarantee safe output.
• No masking or subsetting: teams needing either capability will have to look elsewhere.
• Self-hosted, Python-centered use: this offers deployment control but is not a turnkey hosted application for nontechnical users.

Alternatives

For a broader set of test-data tools, browse the Test Data Management Software category. Windocks is a freemium option with a free plan for a 1 GB database image and one simultaneous container, across Windows and Linux; choose it if those container limits and its no-support free tier match your needs. Data Maestro offers a free plan after a 30-day full trial, then limits exports or database pushes to 100 rows and saved database connections to two; consider it if that trial-led workflow fits better.

MOSTLY AI has a free plan with no further limits stated, making it an option to compare if you want another synthetic-data product. Faker.js is a free, MIT-licensed alternative for commercial and non-commercial use. Mockaroo offers a free plan capped at 1,000 rows per file and 200 API requests per day, with unlimited browser downloads; that may suit smaller browser-based generation needs.

Tonic Fabricate has a free plan with $5 monthly credits, basic data exports and Discord support on Tonic Cloud; consider it if a hosted, credit-based option suits your workflow. Databucket is a free open-source alternative. Synthetic Data Vault offers a free community plan with five data types, five basic constraints, nine models, CSV and Excel support, and local file types; its defined feature set may be preferable when those capabilities meet your needs.

Verdict

Choose Metasyn if you need a free, self-hosted Python workflow for reproducible synthetic tabular data, especially with Polars or pandas and extensible requirements. Its main advantage is the combination of auditable settings, broad data handling and multiple export formats. Look elsewhere if you need masking, subsetting or a hosted, nontechnical workflow—and in any case, review generated data before treating it as safe.