Open Diffix

Free plan APILinuxMacSelf-hostedWindows

What Open Diffix is

Open Diffix is an open-source project for anonymizing structured data. Its SynDiffix Python package generates synthetic datasets intended to be statistically accurate while protecting anonymity. It is especially suited to descriptive analytics such as histograms, heatmaps, column correlations, counts, averages, and standard deviations. Teams can create a separate synthetic table for each analytic task, with the project designed to preserve anonymity across many combinations of columns. SynDiffix can also support machine learning, although the maker describes it as somewhat less easy to use for those applications. A distributable blob lets others generate synthetic tables, but those users need to install SynDiffix; blobs currently do not scale well beyond about 10 columns. SynDiffix installs with pip and requires Python 3.10 or later. For data with multiple rows per protected entity, the entity identifier must be supplied separately to avoid potentially compromising anonymity. The project also includes pg_diffix, a PostgreSQL extension that returns anonymized aggregate answers for a very limited subset of SQL queries. Its development is paused; it requires PostgreSQL 13 or higher and has documented Linux and Windows installation. Open Diffix is free and available for API, Linux, macOS, Windows, and self-hosted use.

Who it suits

Open Diffix suits analysts and data teams that need anonymous synthetic structured data for descriptive analysis, or a way to distribute datasets for others to use. It can also serve machine learning work, though the maker says that use is less easy, and its PostgreSQL extension may suit teams with narrow aggregate-query needs.

What it does well

Generates statistically accurate synthetic data with strong anonymity.
Supports descriptive analytics including counts, averages, and correlations.
Can tailor separate synthetic tables to different analytic tasks.
Open source and free, with API and self-hosted availability.
SynDiffix installs with pip on Python 3.10 or later.

What to know first

Distributable blobs currently do not scale well beyond about 10 columns.
Entity identifiers must be supplied separately for multi-row protected entities.
pg_diffix development is currently paused.
pg_diffix supports only a very limited subset of SQL queries.

Limits

SynDiffix blobs currently do not scale well beyond about 10 columns. pg_diffix only returns aggregate answers for a very limited subset of SQL queries, and its development is paused.

Getting started

Visit the Open Diffix website to explore the project.
Install SynDiffix with pip and use Python 3.10 or later.
For data with multiple rows per protected entity, supply the entity identifier separately.
To use a distributable blob, install SynDiffix first.
For pg_diffix, use PostgreSQL 13 or higher and follow the documented Linux or Windows installation.

Verdict

Choose Open Diffix if you need open-source synthetic data for structured-data analysis, especially descriptive analytics and task-specific datasets. Look elsewhere if you need distributable blobs at scale beyond about 10 columns or a flexible PostgreSQL anonymization extension, since pg_diffix is limited and its development is paused.

Questions

Is Open Diffix free?

Yes. It is listed as free and open source, with a free plan.

What does SynDiffix do?

It creates statistically accurate synthetic data from structured data while aiming to preserve strong anonymity. It is especially suited to descriptive analytics.

Which platforms are supported?

The listed platforms are API, Linux, macOS, self-hosted, and Windows.

What does SynDiffix require?

It installs with pip and requires Python 3.10 or later.

What is pg_diffix?

It is a PostgreSQL extension that returns anonymized aggregate answers for a very limited subset of SQL. It requires PostgreSQL 13 or higher, and its development is paused.

What happens when data has multiple rows per protected entity?

The entity identifier must be supplied separately, or anonymity may be compromised.

Facts from Open Diffix's own pages

Analytics: SynDiffix is especially suited to descriptive analytics, including histograms, heatmaps, column correlations, counts, averages, and standard deviations. · open-diffix.org, 8 Oct 2026
Blob limit: The maker’s repository says SynDiffix blobs currently do not scale well beyond about 10 columns. · github.com, 8 Oct 2026
Distribution: SynDiffix supports creating a distributable blob from which users can generate synthetic tables, and the maker says users need to install SynDiffix to use a blob. · github.com, 8 Oct 2026
Install: The SynDiffix repository says to install the package with pip and requires Python 3.10 or later. · github.com, 8 Oct 2026
Machine learning: SynDiffix can be used for machine learning, though the maker says it is somewhat less easy to use for ML applications. · open-diffix.org, 8 Oct 2026
Origins: The underlying technology for Open Diffix tools was developed by the Max Planck Institute for Software Systems in partnership with the former Aircloak GmbH. · open-diffix.org, 8 Oct 2026
pg_diffix: pg_diffix is a PostgreSQL extension that returns aggregate anonymized answers for a limited subset of SQL queries, and its development is currently paused. · open-diffix.org, 8 Oct 2026
pg_diffix compatibility: The pg_diffix repository says PostgreSQL version 13 or higher is required and documents installation on Linux and Windows. · github.com, 8 Oct 2026
pg_diffix limitations: The Open Diffix site says pg_diffix only returns aggregate answers consisting of column values and associated counts for a very limited subset of SQL, and is less flexible than SynDiffix. · open-diffix.org, 8 Oct 2026
pg_diffix requirements: The pg_diffix repository requires PostgreSQL 13 or higher and documents installation for Linux and Windows. · github.com, 8 Oct 2026
pg_diffix support: The maker says pg_diffix development is currently paused and invites interested project teams to contact them. · open-diffix.org, 8 Oct 2026
PostgreSQL extension: pg_diffix is a PostgreSQL extension that returns aggregate anonymized answers to a very limited subset of SQL queries. · open-diffix.org, 8 Oct 2026
Project: Open Diffix is an open-source project developing tools for anonymizing structured data. · open-diffix.org, 8 Oct 2026
Protected entities: For data with multiple rows per protected entity, the entity identifier must be supplied separately or anonymity may be compromised. · github.com, 8 Oct 2026
Runtime requirement: SynDiffix installs with pip and requires Python 3.10 or later. · github.com, 8 Oct 2026
Support: The maker invites questions and requests for help or advice by email and bug reports or feature requests through GitHub issues. · open-diffix.org, 8 Oct 2026
SynDiffix: SynDiffix is a Python package for generating statistically accurate and strongly anonymous synthetic data from structured data. · open-diffix.org, 8 Oct 2026
Tailored datasets: SynDiffix is designed to generate a separate synthetic dataset tailored to each use case, and the site says it can support thousands of column combinations while preserving anonymity. · open-diffix.org, 8 Oct 2026
Usage: SynDiffix can create separate synthetic tables tailored to individual analytic tasks while maintaining strong anonymity across many column combinations. · open-diffix.org, 8 Oct 2026

Where Open Diffix stands on Freedom251

Data Masking Software#9 of 26