Synth Studio

Synth Studio is a tabular synthetic-data generator for teams working with sensitive records. It suits healthcare, fintech, machine-learning and enterprise data-sharing workflows. Its combination of configurable differential privacy, quality evaluation and deployment flexibility makes it worth considering, especially for teams that can work around its current API authentication requirements.

Overview

Import CSV or JSON files and Synth Studio detects their schemas automatically. It can generate synthetic datasets using CTGAN, TVAE, GaussianCopula or schema-based methods, giving teams multiple approaches to tabular data rather than a single generation path. The tabular focus is a clear boundary: it is not aimed at other data modalities.

Before generation, the software can detect sensitive fields such as names, email addresses and SSNs. Differential privacy uses configurable epsilon and delta settings with RDP accounting, while evaluation tools assess statistical similarity, machine-learning utility and privacy leakage. Together, these controls help teams scrutinize both data usefulness and exposure risk; they do not remove the need to judge whether a particular generated dataset suits its intended use.

Synth Studio is open source and can run self-hosted or through its cloud offering. Its architecture uses PostgreSQL for metadata, Redis for queueing and cache, and AWS S3 for files. That deployment choice is useful for organizations with different data-handling needs, though teams planning integrations should account for the current API access hurdle: direct access requires a session cookie or trusted proxy headers, with API keys upcoming.

Key features

Privacy and governance

Compliance reporting covers HIPAA, GDPR, CCPA and SOC-2, with audit logs and model cards. Secure authentication and access control, privacy-budget enforcement, audit trails and lineage add governance controls around generation. These features make Synth Studio a stronger fit for teams that need to track and manage sensitive-data workflows than for users seeking only a basic generator.

API and integrations

The REST API guide gives limits of 1,000 authenticated requests per minute, 10 file uploads per minute, five concurrent generation jobs and 5,000 API calls per hour. Those ceilings provide defined operating limits, but the session-cookie or trusted-proxy requirement may complicate automated integrations until API keys arrive. Support resources include documentation, GitHub Issues, GitHub Discussions and a security contact email.

Pricing

Synth Studio's Free to Use plan costs nothing and allows up to 1M rows per generation under an MIT License. The row allowance makes it a substantial starting point for teams evaluating synthetic tabular data or running workloads within that cap, while the free price avoids an entry fee. The stated plan terms do not specify seat counts, a trial period or renewal conditions.

Platforms

Synth Studio supports API, Linux, macOS, self-hosted, web and Windows use. Local installation documentation covers Windows 10+, macOS 10.15+ and Linux. The choice of cloud or self-hosted deployment is useful for teams balancing convenience against control over where the software runs.

Who it's for

Healthcare and fintech teams handling sensitive records are natural candidates, as are machine-learning groups that need synthetic tabular data and evaluation of its utility. Enterprise data-sharing teams may value the privacy controls, compliance reporting and self-hosting option. It is less suitable for projects requiring non-tabular data or integrations that depend on API keys today.

Pros and cons

• Pros: Four generation methods and automatic CSV/JSON schema detection give teams flexibility within tabular data workflows.
• Pros: Configurable differential privacy, quality checks and sensitive-field detection address privacy and usefulness together.
• Pros: Open-source self-hosting and a cloud option let organizations choose a deployment approach.
• Cons: Direct API access depends on session cookies or trusted proxy headers, which can make automated integration less straightforward.
• Cons: Its data scope is tabular, ruling it out for other modalities.

Alternatives

AI Synthetic Data Generators and Synthetic Data Generation Software are useful starting points for comparing options across the category.

Choose REaLTabFormer instead if a free, MIT-licensed option for Linux, macOS, Windows or self-hosting meets the need and Python 3.7 or later is acceptable. MOSTLY AI is another free-start option for teams looking across API, Linux, self-hosted and web deployments.

Synthehol Dataset may suit a small project that can work within its Free plan's 1,000 monthly credits, 50K rows per month, five projects, five datasets, one user, 1 GB storage and 30-day retention. For a cloud-oriented option with basic exports and Discord support, consider Tonic Fabricate, whose Free plan includes $5 monthly credits.

SynthCity is a free open-source Python library for Linux, while DataSynthesizer is a free MIT-licensed option for self-hosted or web use that installs with pip and requires first-normal-form table input. SimpleTest is a freemium API, extension and web option with a Starter plan allowing 50 parallel test runs and 500 execution minutes per month, with AI-generated datasets limited to one per test. NVIDIA ShadowPlay is a free Windows option requiring Windows 10 or 11, a GeForce 551.52 driver or later and a supported GPU.

Verdict

Synth Studio is a strong candidate for teams generating sensitive tabular data that need privacy settings, evaluation and a choice between self-hosting and cloud deployment. Its free plan and governance controls are compelling reasons to start with it; look elsewhere if the project needs other data modalities or direct API-key access now.