Synth Studio

Synth Studio generates synthetic tabular datasets for organizations working with sensitive data. It best suits healthcare, fintech, machine-learning and enterprise data-sharing teams that need privacy controls and deployment flexibility. Its broad privacy and evaluation toolkit is a strength; teams planning API-driven workflows must account for authentication requirements that currently rule out API keys.

Overview

Import CSV or JSON and Synth Studio detects the schema automatically. It can generate data with CTGAN, TVAE, GaussianCopula or schema-based methods, giving teams several approaches for tabular datasets. The focus is strictly tabular, so it is not the choice for work involving other data modalities.

Automatic detection of fields such as names, emails and SSNs helps surface sensitive information before generation. Differential privacy offers configurable epsilon and delta settings with RDP accounting, while evaluation tools measure statistical similarity, machine-learning utility and privacy leakage. Those checks help teams weigh usefulness against exposure rather than treating synthetic output as inherently safe or fit for purpose.

Synth Studio is open source and supports self-hosting as well as cloud deployment. Local installation documentation covers Windows 10+, macOS 10.15+ and Linux. Its architecture uses PostgreSQL for metadata, Redis for queueing and cache, and AWS S3 for files, which matters to teams planning an installation around those components.

Key features

Privacy and governance

The product combines privacy-budget enforcement, secure authentication and access control with audit trails and lineage. Compliance reporting covers HIPAA, GDPR, CCPA and SOC-2, with audit logs and model cards. This is a substantial governance-oriented package for teams that need to document data-generation workflows, though the presence of reporting does not by itself establish compliance for a particular organization or use.

API access and limits

REST API integration supports automated workflows, with documented limits of 1,000 authenticated requests per minute, 10 file uploads per minute, five concurrent generation jobs and 5,000 API calls per hour. Direct access currently requires a session cookie or trusted proxy headers; API keys are upcoming. That makes the API less straightforward for teams expecting conventional key-based service authentication. Support resources include documentation, GitHub Issues, GitHub Discussions and a security contact email.

Pricing

Synth Studio has one published plan: Free to Use: free. It allows up to 1M rows per generation and is MIT-licensed. There is no paid tier described, so teams needing a larger per-generation allowance should not assume an upgrade path or expanded quota. The free plan is a strong fit for evaluation and workloads within the stated cap; the row limit is the key constraint for larger synthetic datasets.

Platforms

Synth Studio is available through API, web and self-hosted deployments, with support for Windows, macOS and Linux. The combination suits teams choosing between hosted access and local installation, while the documented API authentication requirements may complicate integration-led deployments.

Who it's for

Healthcare and fintech teams, machine-learning groups and organizations sharing data across enterprise boundaries are the clearest fit. They can use configurable differential privacy, quality assessment, sensitive-field detection and compliance reporting in one tabular-data workflow. It is less suitable for teams needing non-tabular generation, a larger stated per-generation allowance, or direct API access through API keys today.

Pros and cons

• Pros: Configurable differential privacy with RDP accounting gives privacy-conscious teams explicit parameters to work with.
• Pros: Statistical, utility and leakage evaluation helps teams judge generated data on more than output volume alone.
• Pros: Open-source self-hosting and cloud deployment offer flexibility for different operating environments.
• Pros: Automatic PII detection and compliance reporting support sensitive-data governance workflows.
• Cons: Generation is limited to tabular data, excluding teams that need other modalities.
• Cons: The free plan caps each generation at 1M rows, a material constraint for larger workloads.
• Cons: API access relies on session cookies or trusted proxy headers for now, rather than API keys.

Alternatives

REaLTabFormer is a free, MIT-licensed option for teams comfortable with Python 3.7 or later and Windows, macOS, Linux or self-hosted deployment.

MOSTLY AI is worth considering for teams looking for a freemium option across API, web, Linux and self-hosted platforms; its free plan has no further limits stated.

NVIDIA ShadowPlay is another free option.

SimpleTest is a freemium API, extension and web tool whose Starter plan allows 50 parallel test runs and 500 execution minutes per month, with AI-generated datasets limited to one per test.

SynthCity is a free, open-source Python library installed from PyPI or source, suited to Linux-based library workflows.

Synthehol Dataset is a free API, self-hosted and web option with 1,000 credits and 50K rows per month, plus five projects, five datasets, one user, 1 GB storage and 30-day retention.

Tonic Fabricate offers a cloud free plan with $5 monthly credits and basic data exports, alongside a Plus plan at $29 a month with $25 monthly credits and metered additional usage.

DataSynthesizer is a free MIT-licensed option installable with pip; its input must be a first-normal-form table.

Browse AI Synthetic Data Generators or Synthetic Data Generation Software for more options.

Verdict

Synth Studio is a compelling free choice for teams generating tabular data from sensitive records, particularly when configurable differential privacy, quality evaluation and self-hosting matter. Its 1M-row generation cap and current API authentication requirements are the clearest reasons to look elsewhere.