SCheckOps
Synthetic Data Vault (SDV)
The Synthetic Data Vault (SDV) is a Python library from DataCebo that learns patterns from real tabular data with machine learning and generates synthetic data, with evaluation reports; it is source-available under the Business Source License.
Latest public signals
Not refreshed yet. Refresh reads the public site and, when the entry has a GitHub repository, the latest release.
Introduction
The Synthetic Data Vault is a Python library designed to be a one-stop shop for creating tabular synthetic data. It uses a variety of machine learning algorithms to learn patterns from real data and emulate them in synthetic data, according to its README. It is part of the SDV project from DataCebo.
SDV supports single tables, multiple related tables and sequential data. Synthesizers include statistical models such as Gaussian copulas and deep learning models such as CTGAN. Metadata describes columns and relationships, constraints enforce business rules, and evaluation tools compare synthetic and real data for quality and privacy.
The licence is the Business Source License 1.1, which allows free use with restrictions rather than being OSI open source; DataCebo sells SDV Enterprise. Release 2.2.0 shipped in November 2026.
In the quality loop it supports realistic test and analytics data without exposing real records.
Features
- ML synthesizers: Gaussian copula, CTGAN and more.
- Multi-table: Preserve relationships.
- Sequential data: Time series support.
- Constraints: Enforce business rules.
- Evaluation: Quality and privacy reports.
Pros
- Statistical realism: Mimics real distributions.
Cons
- Source-available licence: Business Source License 1.1, not OSI open source.
- Needs real data: Training requires source data.
Usage and Links (Example)
from sdv.single_table import GaussianCopulaSynthesizer
synth = GaussianCopulaSynthesizer(metadata)
synth.fit(real_df)
synthetic = synth.sample(num_rows=1000)Resources:
Pricing
Free under the Business Source License 1.1 with restrictions; SDV Enterprise is commercial. See site.
Recommended for?
- Data and QA teams needing statistically realistic data.
- Not a fit if you require an OSI licence.
Conclusion
SDV is a powerful synthetic data library; check its licence terms for your use.
On CheckOps, Synthetic Data Vault (SDV) is filed under the Integration testing stage. Related tools listed here include MOSTLY AI, Synthea, Tonic Structural, GenRocket, Faker (Python) and Great Expectations (GX). Before choosing, compare them on licence, supported languages and platforms, hosting model, and how well each fits your existing CI pipeline. Facts on this page were checked against the official website, documentation and repository on 11 November 2026; statements the vendor makes about its own product are marked as vendor claims.
Alternatives
Other tools on CheckOps that cover similar ground. Open a page, or compare one with Synthetic Data Vault (SDV) side by side.
MMOSTLY AIMOSTLY AI offers an Apache-2.0 open-source Synthetic Data SDK for training generators on tabular or language data and creating privacy-safe synthetic datasets locally, plus a commercial synthetic data platform.Compare with Synthetic Data Vault (SDV)
SSyntheaSynthea is an Apache-2.0 synthetic patient population simulator that generates realistic but not real patient records in HL7 FHIR, C-CDA, CSV and other formats for healthcare software testing.Compare with Synthetic Data Vault (SDV)
TTonic StructuralTonic Structural is a commercial test data management product from Tonic.ai that de-identifies, subsets and synthesizes structured and semi-structured data from production into safe, production-like test data.Compare with Synthetic Data Vault (SDV)
GGenRocketGenRocket is a commercial enterprise synthetic test data platform built on 'design-driven' data, generating data from rules and scenarios without copying production, with self-service, CI/CD integration and masking options.Compare with Synthetic Data Vault (SDV)
FFaker (Python)Faker is an MIT-licensed Python package that generates fake but realistic data such as names, addresses, text, dates and identifiers, with dozens of locales, a pytest fixture and seeding for reproducible test data.Compare with Synthetic Data Vault (SDV)
GGreat Expectations (GX)Great Expectations (GX Core) is an Apache-2.0 Python framework for data quality testing that lets teams write Expectations, expressive unit tests for data, validate datasets and generate Data Docs; GX Cloud adds a hosted service with a free Developer option.Compare with Synthetic Data Vault (SDV)