If you mainly use Tonic Structural to mask and subset production databases, the enterprise alternatives are Perforce Delphix (masking plus virtual, writable database copies), K2view (test data organised by business entity across many systems), DATPROF (modular masking, subsetting and a self-service portal) and, in IBM or Informatica estates, IBM Optim and Informatica Test Data Management. If you want synthetic data designed per test case rather than copied from production, look at GenRocket. If you use Tonic Fabricate to fill development databases with fake rows, Mockaroo, Redgate SQL Data Generator and dbForge Data Generator publish their prices and cost far less, but none of them reads or masks production data.
Tonic.ai sells three test data products. Tonic Structural reads a source database, applies a generator to each sensitive column (masking, consistent replacement or synthesis), optionally subsets the data and writes a de-identified copy. Tonic Fabricate generates relational data, free text and mock APIs from a description or schema; Tonic acquired it in April 2025, describing it as a product built by Mockaroo. Tonic Textual redacts or synthesizes sensitive values in free text and files. Structural connectors cover the main relational engines, cloud warehouses, MongoDB and Salesforce.
Teams that look elsewhere usually give one of these reasons; the first three come from Tonic's own pricing page and documentation, the last is our editorial reading.
- Structural is priced by source data size, by quote. Professional and Enterprise are annual contracts based on on-disk source size, with no published rates.
- Self-hosting needs Enterprise. The Professional plan runs only on Tonic Cloud (up to 10 TB, 10 users and 2 source types); teams that must keep data on their own infrastructure need the Enterprise plan.
- Subsetting is not available everywhere. Tonic documents that some connectors, including Databricks, BigQuery, Iceberg, files and Spark, do not support subsetting.
- Three products, three pricing models (editorial). Data-size contracts for Structural, credits for Fabricate and per-word billing for Textual make total cost harder to estimate than a single licence.
For head-to-head pages, see Tonic.ai vs GenRocket, Delphix vs K2view and Redgate SQL Data Generator vs dbForge Data Generator.
Quick picks
Virtual, writable database copies plus masking that keeps referential integrity across sources.
Subsets, masks and reserves complete business entities, such as one customer, across every system that holds them.
Separate masking, subsetting and self-service modules with a licence that does not depend on database size.
Test data, privacy and archiving across Db2 for z/OS, IMS, VSAM and distributed databases.
Model-driven generation with generators and receivers for databases, files and APIs.
Browser-based generator with a free plan and annual plans from USD 60.
How we chose
We considered Perforce Delphix, K2view, DATPROF, IBM Optim, Informatica Test Data Management, GenRocket, Mockaroo, Redgate SQL Data Generator, dbForge Data Generator and QuerySurge. To be included, a product had to be sold in November 2026 and either mask or subset copies of production data, or generate synthetic data for databases, with its pricing model or licence stated on an official page. QuerySurge was left out: it validates data by comparing source and target queries, but does not generate, mask or subset test data. Informatica Test Data Management is included with a caveat, because its current cloud edition is scoped to Salesforce orgs and its on-premises product depends on PowerCenter.
For each product we recorded its approach (masking and subsetting, virtualisation, entity-based provisioning or synthetic generation), its pricing model, where it runs, the data sources it names and its main limitation. Most enterprise test data vendors sell by quote, so prices are quoted only where the vendor publishes them. None of this comes from our own use of the tools.
The order is not a ranking. Entries start with enterprise test data management platforms, then model-driven synthetic data, then lightweight generators.
At a glance
| Tool | Approach | Pricing | Deployment | Data sources | Main trade-off |
|---|---|---|---|---|---|
| Perforce Delphix | Virtual database copies plus masking; synthetic module | Quote, by capability and deployment size | On premises; AWS, Azure, Google Cloud, OCI, IBM Cloud | Oracle, SQL Server, Db2, SAP HANA, PostgreSQL, MongoDB, Snowflake and more | A platform to deploy and operate |
| K2view | Entity-based subsetting, masking and synthetic data | Quote; Start Free offer with unstated terms | On premises, cloud, hybrid | Relational, NoSQL, mainframe, SaaS, files | Entity model to design up front |
| DATPROF | Masking, subsetting, self-service portal | Quote; fixed modular licence | On premises or self-managed cloud | Oracle, SQL Server, Db2, PostgreSQL, MySQL, MariaDB | No cloud warehouse or NoSQL support listed |
| IBM Optim | Test data, masking and archiving by business object | Quote via IBM | Containers on premises or cloud; z/OS via zCX | Db2 (distributed and z/OS), Oracle, IMS, VSAM | Optim 2.0 is new (July 2026) |
| Informatica TDM | Discovery, subsetting, masking, generation | Quote; IDMC masking metered in IPUs | On premises (10.5.x) or IDMC | PowerCenter connectivity; Salesforce in the cloud edition | Cloud edition is Salesforce-only |
| GenRocket | Model-driven synthetic data, plus subsetting and masking | Annual per-project licence, 20-project minimum, quote | Hosted platform; runtime on your machines | JDBC databases, VSAM, XML, JSON, CSV, EDI, Salesforce | Upfront modelling work |
| Mockaroo | Schema-based fake data and mock APIs | Free; Silver USD 60 and Gold USD 500 a year | Web; Docker on Enterprise | Exports SQL, CSV, JSON, Excel, XML | Cannot read or mask production data |
| Redgate SQL Data Generator | Synthetic data from a SQL Server schema | USD 303 per user a year (1-4 users) | Windows | SQL Server, Azure SQL Database | SQL Server only |
| dbForge Data Generator | Synthetic data from a SQL Server schema | From USD 199.95 a year | Windows | SQL Server, Azure SQL, Fabric, RDS | One engine per product |
Perforce Delphix
Delphix, now sold by Perforce as Perforce Delphix, does two things Tonic Structural approaches differently. Continuous Data keeps a compressed record of source databases and delivers virtual copies that share common blocks, so each copy is a full read/write database that can be refreshed, rewound, bookmarked and branched without the storage of a physical copy. Continuous Compliance discovers sensitive data and masks it with realistic values that keep referential integrity within and across sources, and Hyperscale Compliance runs masking in parallel for very large datasets, including Snowflake and Databricks. A Synthetic Data module covers edge cases.
Perforce publishes no prices; its pricing page says the price depends on the capabilities you choose and the deployment size, without metered billing. In our view Delphix suits teams whose pain is the number and size of test environments, not only masking. See Delphix vs K2view.
- No public pricing.
- A platform with engines to deploy and operate, not a lightweight tool.
- Managed cloud databases such as Amazon RDS and Azure SQL Database get API orchestration rather than virtual copies.
K2view
K2view organises test data by business entity, such as a customer, account or order, stored in its own Micro-Database on the K2view Data Product Platform. Instead of following table joins in one database, it subsets by business attributes (for example account status or balance) and returns the complete data for each selected entity from every system that holds it. Teams provision data through a self-service portal, API or agent, reserve entities so tests do not overwrite each other, and roll back to a snapshot. Masking can run statically or in flight and keeps referential integrity at entity level, and synthetic data can be generated with AI, rules or entity cloning.
K2view lists relational and NoSQL databases, mainframe data, SaaS applications, files and cloud data stores as sources. It publishes no prices; the site offers a Start Free option whose terms were not stated on the pages checked. See Delphix vs K2view.
- No public pricing.
- The entity model must be designed up front, which takes effort.
- No database-by-database support matrix on the public product pages.
DATPROF
DATPROF is a Dutch test data management suite split into modules: Privacy masks data and generates synthetic values (over 50 built-in generators), Subset extracts a smaller, consistent slice of production, Runtime is a browser portal and REST API that deploys and schedules the masking and subset templates, and Analyze and Virtualize add discovery and virtual copies. Native support covers Oracle, SQL Server, Db2 for i, Db2 LUW, PostgreSQL, EDB, MySQL and MariaDB, plus JDBC sources and files.
Prices are by quote, but DATPROF says the licence is a fixed price that does not depend on database size, which is the main contrast with Tonic Structural's data-size pricing.
- No published prices; pricing needs a sales form.
- Each capability is a separate module, so a full setup means several licences.
- Cloud warehouses and NoSQL databases are not on the native support list.
IBM Optim
IBM Optim (formerly IBM InfoSphere Optim) combines Optim Test Data Management, Optim Data Privacy and Optim Archive. Optim 2.0, generally available on 31 July 2026, is a containerised application with an Apache Spark runtime that treats a business object, a set of related rows across systems, as the unit of work, so one customer ID can be masked to the same value in Oracle and Db2. It runs on premises, in private or public cloud, and natively on z/OS through z/OS Container Extensions, with Db2, Db2 for z/OS, Oracle, IMS and VSAM among the named sources.
IBM publishes no prices and no trial on the pages checked. Some older InfoSphere Optim SKUs are being withdrawn, so check the lifecycle page for your product ID.
- No public prices or trial; purchase runs through IBM sales.
- Optim 2.0 is a new release, so moving from InfoSphere Optim needs planning.
- IBM's product pages do not list every supported database.
Informatica Test Data Management
On premises, Informatica TDM works with PowerCenter and PowerExchange: it profiles sensitive fields and key relationships, builds subsets from related tables, applies masking rules such as substitution, shuffle and key masking, and generates data from rules. In the cloud, Informatica's October 2026 product schedule lists Cloud Test Data Management Standard Edition, which masks and subsets a Salesforce sandbox org from a production org; masking of other sources in IDMC is a separate Data Masking capability metered in IPUs.
Salesforce completed its acquisition of Informatica on 18 November 2025. Support for TDM on the 10.4 platform ended on 31 March 2024, and the public roadmap is thin, so confirm plans with Informatica before committing (editorial).
- The packaged cloud edition is Salesforce-only; other sources need IPU-metered masking in mappings.
- On-premises TDM depends on the PowerCenter platform.
- No published prices for any option.
Read our Informatica Test Data Management review · Official site
GenRocket
GenRocket generates test data from a model instead of a copy of production. A Domain is modelled like a table and its Attributes like columns; each attribute gets a Generator (GenRocket cites 750+), and Receivers write output to JDBC databases, files such as XML, JSON, CSV and VSAM, web services or Salesforce. Scenarios set row counts and keep parent and child records consistent. Subsetting and masking are available too, for Oracle, Db2, SQL Server, MySQL, PostgreSQL and Sybase, along with an API for test frameworks and CI/CD pipelines.
GenRocket sells an annual licence per test data project with a minimum of 20 projects and unlimited users; every price is by quote. It is the closest alternative to Tonic Fabricate for teams that need synthetic data at enterprise scale (editorial). See Tonic.ai vs GenRocket.
- Minimum purchase of 20 test data projects, with all prices by quote.
- Modelling domains and generators takes upfront design work compared with masking an existing copy.
- Detailed support matrices sit in a knowledge base that requires a login.
Mockaroo
Mockaroo lets you define a schema in the browser, pick a data type for each field and download rows as CSV, JSON, SQL INSERT statements, Excel or XML, with a REST API and mock API endpoints. Its creator later built Fabricate, which Tonic.ai acquired in 2025, but Mockaroo itself continues as a separate service under Mockaroo, LLC.
The Free plan allows up to 1,000 rows per file and 200 API requests a day on the pricing page, though the API documentation states a 5,000-record download limit for free accounts. Silver is USD 60 a year, Gold USD 500 and Enterprise USD 7,500 with a self-hosted Docker image. It replaces Fabricate for simple fixtures, not Structural.
- Does not read production databases, so it cannot mask or subset real data.
- Free plan limits are small.
- The pricing page and API docs state different free download limits.
Redgate SQL Data Generator
SQL Data Generator reads a SQL Server schema, proposes a generator for each column from names, types and constraints, and inserts the rows while following foreign keys. It runs as a desktop tool, inside SSMS and from the command line. Version 4.8.0 added SSMS 22 and SQL Server 2025 support with a documented known issue, and 4.8.16 shipped on 5 November 2026.
It is sold standalone at USD 303 per user a year (1-4 users) and inside SQL Toolbelt Essentials. See Redgate SQL Data Generator vs dbForge Data Generator and Redgate alternatives.
- SQL Server only, and Windows only.
- Creates synthetic data; it does not mask or subset production databases.
- Redgate's requirements and product pages list older SQL Server versions than the release notes.
dbForge Data Generator for SQL Server
Devart's dbForge Data Generator fills SQL Server tables with synthetic data through a Data Population Wizard, and can run the generated script, save it or open it for review. Devart documents support from SQL Server 2000 (limited) to SQL Server 2025, SQL Server on Linux, Azure SQL Database and Managed Instance, Azure Synapse, Microsoft Fabric and Amazon RDS. It is also available as an SSMS add-in and inside dbForge Studio for SQL Server.
The product page lists it from USD 199.95 a year. Devart sells separate generators for MySQL and MariaDB, Oracle and PostgreSQL. See Redgate SQL Data Generator vs dbForge Data Generator.
- Windows only.
- One engine per product: other databases need separate Devart generators.
- Generates synthetic data; no masking or subsetting of production copies.
Read our dbForge Data Generator for SQL Server review · Official site
When to stay with Tonic.ai
Stay with Tonic if you want de-identified copies of relational databases, cloud warehouses, MongoDB and Salesforce from a product you can start on Tonic Cloud without deploying your own platform, with an API to automate refreshes (editorial). Tonic is also one of the few vendors here covering structured data, synthetic generation and free text redaction (Textual) under one roof, and Fabricate's Free and Plus plans publish prices (editorial).
Switching makes most sense when the need is narrower or different: space-efficient, refreshable copies of very large databases (Delphix), consistent test data across many applications (K2view), a fixed licence on a relational estate (DATPROF), mainframe sources (IBM Optim), model-driven synthesis (GenRocket) or simple fake data for SQL Server development (Redgate, dbForge, Mockaroo). Ask each vendor to run a proof of concept on a masked sample of your own schema before you sign.
Frequently asked questions
What is the main difference between Tonic.ai and Delphix?
Both mask sensitive data in copies of production. Delphix also delivers virtual, writable database copies that share storage and can be refreshed or rewound, while Tonic Structural writes a de-identified physical copy and adds subsetting and synthesis. See Delphix vs K2view for how Delphix compares with an entity-based approach.
Is there a free alternative to Tonic.ai?
For synthetic fake data, Mockaroo has a free plan, as does Tonic Fabricate itself. For masking and subsetting production databases, none of the tools on this page publishes a free edition; DATPROF and K2view offer a test drive or a Start Free option whose terms were not stated on the pages we checked.
Is Mockaroo part of Tonic.ai?
No. Tonic.ai acquired Fabricate, a separate synthetic data product built by Mockaroo, in April 2025; the announcement does not say Mockaroo itself was acquired. Mockaroo continues as its own service under Mockaroo, LLC.
Which alternative is closest to Tonic Fabricate?
GenRocket is the enterprise option for synthetic data generated from a model. For smaller jobs, Mockaroo, Redgate SQL Data Generator and dbForge Data Generator create fake rows at published prices. See Tonic.ai vs GenRocket.
Sources
- Tonic.ai pricing
- Tonic Structural data connector summary
- Tonic Structural self-hosted deployment
- Tonic.ai acquires Fabricate
- Perforce Delphix
- Perforce Delphix pricing
- K2view test data management
- DATPROF products
- DATPROF FAQ
- IBM Optim 2.0 announcement
- Informatica Cloud and Product Description Schedule
- GenRocket pricing
- Mockaroo pricing
- Redgate SQL Data Generator
- dbForge Data Generator for SQL Server
Checked November 2026.
How we research these guides: our editorial method.