testdatatools
AI & ML datasets

SDV Community / Enterprise

SDV learns patterns from existing tables and samples synthetic data through Python. Community includes single-table, multi-table and sequential workflows. Enterprise capabilities and license terms are separate. 1

Sources reviewed

Manufacturer / project: DataCebo / SDV project

When does it fit?

Consider SDV when you can train on representative tables and measure the patterns your tests need. It offers a local workflow for statistical data generation and evaluation. 1 · 2

Fit still depends on your mandatory requirements.

What should you check before choosing?

For a fixed payment reversal or exact regression result, use explicit cases or verify the required constraints. Community HMA targets smaller relational graphs and does not document conditional sampling. A saved Python model is not a portable model for every engine. 1 · 2 · 3 · 4 · 5 · 6

What must your PoC prove?

Check foreign keys, constraints, rare cases and source/output overlap. Keep evaluation data out of training. Reload the saved model on pinned versions and test reset_sampling. Confirm that the intended commercial use is permitted.

Acceptance plan, not a product test result. Execution scope and limits.

Deployment and license

Community runs locally in Python on CPU. Your team manages dependencies and storage. Release 2.1.0, 30 October 2026, removes Python 3.9 support; 2.0 changes metadata and sampling APIs. Test migration of existing code. 1 · 2 · 3 · 4 · 5 · 6

Community is source-available under BSL 1.1, not blanket open source. Check the additional-use restriction and terms of the installed version. Enterprise requires a separate license. 1 · 2 · 3 · 4 · 5 · 6

Confirm the entitled version and operating requirements.

How you work with it

Fit a synthesizer to source tables, sample dataframes and validate them. Save/load supports reuse in compatible SDV/Python environments. Treat Python pickle files as trusted executable inputs. 1 · 2 · 3 · 4 · 5 · 6

Export does not establish compatibility with another runtime.

Documented capabilities

De-identification / maskingUnclear
Subsetting / subset planningUnclear
Data virtualizationUnclear

Documentation is not a measured benchmark. Unclear is missing evidence; the comparison matrix explains the gap.

Sources and scope

Sources can cover different releases; they do not establish measured performance.

Open the cited sources (6)
  1. SDV Community

    Local Python library for single-table, multi-table and sequential learned data; CPU operation.

    Source checked: 2026-10-11
  2. SDV Business Source License 1.1

    Source-available terms with additional-use restriction on commercial synthetic-data services to third parties; version-specific conversion terms.

    Source checked: 2026-10-11
  3. HMA multi-table synthesizer

    Community HMA guidance for small relational graphs; conditional sampling not supported in the reviewed HMA docs; Enterprise is separate.

    Source checked: 2026-10-11
  4. Sampling and reset_sampling

    Repeated sampling of the same fitted model; not identical training or cross-platform byte replay.

    Source checked: 2026-10-11
  5. GaussianCopula fitting and model save/load

    Statistical single-table fit, pandas data and local saved Python model; not vendor-neutral engine interchange.

    Source checked: 2026-10-11
  6. SDV Community release notes

    2.0.0 (15 October 2026) changes metadata and sampling APIs; 2.1.0 (30 October) removes Python 3.9. Release notes, not a hands-on migration test.

    Source checked: 2026-11-13

Use the same acceptance checks for your shortlist