DataSynthesizer review
A free local choice for privacy-aware synthetic tabular data generation.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
DataSynthesizer is an open-source Python library with a local web interface for generating synthetic datasets from sensitive tabular data. It is aimed at data scientists and data owners who need to share structurally and statistically similar records without releasing the original dataset. The project supports self-hosted deployment and offers a local Django-based web UI alongside Python installation through pip, making it suited to teams that want to keep generation within their own environment.
Its main strength is the range of tabular generation approaches. Random mode can create type-consistent synthetic records, while independent-attribute mode models attributes separately. Correlated-attribute mode uses a differentially private Bayesian network to model relationships between attributes while applying privacy-preserving techniques. DataSynthesizer can automatically infer attribute types and distributions, and it supports numerical, categorical, string, and datetime attributes. Synthetic datasets can be generated at arbitrary sizes, which suits workflows that need larger development or sharing datasets from a sensitive source.
The product remains deliberately narrow in scope. Its input model assumes a single table in first normal form, so teams working with relational structures should look for a generator designed around verified multi-table relationships. It also focuses on tabular records rather than unstructured data such as documents or media. DataSynthesizer is a strong fit for privacy-conscious teams seeking a free, local tool for single-table tabular synthesis, particularly when Python or a local web interface fits the workflow. Teams needing relational or unstructured generation should choose a broader alternative.
DataSynthesizer pros and cons
- Where it wins
- Free and open source with local deployment options
- Supports differential privacy and Bayesian-network modeling
- Handles numerical, categorical, string, and datetime attributes
- Where it doesn't
- Focused on tabular data rather than unstructured data
- Does not support verified relational data generation
- Requires Python installation or a local Django-based web UI
DataSynthesizer fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.
Last updated · How we research and update