This guide provides details about our pricing SDV Enterprise’s payment structure combines a subscription fee with flexible, usage-based pricing, both paid on a monthly basis. Our pricing model is designed to be transparent and predictable, so you can pay based on your actual usage and maintain full control through configurable spending caps. This guide explains how it all works and answers common questions about subscriptions, bundles, spending caps, and billing.

​

Benefits of Pay-As-You-Go Pricing

• No annual subscription commitments.
• No large upfront purchases.
• Pay only for the synthetic data you generate.
• Cost scales with your usage needs.

​

Subscription costs

Subscriptions are charged on a recurring monthly basis. To calculate your flat monthly subscription cost, combine the cost of SDV Enterprise Base with the cost of any additional bundles you need.

​

Usage-based pricing and costs

Your subscription provides access to SDV Enterprise Base, so that you can build and train generative models. Product usage is measured when you generate synthetic data from those trained models. The unit of usage is a sampled cell. A sampled cell represents a single generated data value within a table. For a single table, the total number of sampled cells is calculated as: Number of rows generated × Number of columns in the table. For example, if you generate a table with 100,000 rows and 20 columns, that’s a total of 2,000,000 sampled cells. Our current pricing is $1 per million sampled cells. The table below shows estimated costs for various numbers of sampled cells.

​

Estimating sampled cell counts for multi-table datasets

For multi-table datasets, the total number of sampled cells is the sum of the sampled cells generated across all tables in the dataset. To estimate usage for a multi-table model, use the scale parameter provided by SDV. The scale parameter controls the overall size of the synthetic dataset relative to the original data used to train the model.

• scale = 1.0 generates approximately the same amount of data as the original dataset.
• scale = 0.5 generates approximately half as much data.
• scale = 2.0 generates approximately twice as much data.

When sampling from a multi-table model, you don’t specify the number of rows for each table individually. Instead, SDV automatically generates data across all tables, while preserving the learned parent-child relationships and cardinality distributions. As a result, the number of rows produced in each table is determined by the model. Because the generated data volume is approximately proportional to the scale factor, a simple way to estimate usage is: Estimated Sampled Cells ≈ Scale × Total Data Cells in the Original Training Dataset Where the total number of data cells in the original dataset is the sum of: Rows × Columns across all tables used to train the model. For example, if your original multi-table dataset contains 500 million data cells, and you sample with scale = 2.0, you can expect to generate approximately 1 billion sampled cells.

​

Purchase and subscribe via DataCebo portal

You’ll manage your SDV Enterprise account through the DataCebo Portal. There you can purchase or cancel subscriptions, manage bundles, monitor usage, and access license keys. If you haven’t already logged in, get started today by visiting this link. To start building with SDV Enterprise, you’ll follow these three steps.

​

1. Subscribe to SDV Enterprise Base.

Once you have logged into the DataCebo Portal, you can subscribe to SDV Enterprise Base by navigating to the Subscriptions page and adding the subscription to your account.

​

2. (Optional) Subscribe to additional bundles.

Once SDV Enterprise Base has been added, you can select and subscribe to additional bundles, such as AIConnectors, XSynthesizers, or Targeted Sampling. Once you have completed your purchase, your subscription will be shown as “Active.” Your license key will allow you to download and install the package, so you can start building.

​

3. Set a monthly spending cap for usage-based synthetic data generation.

Usage-based charges are controlled through a monthly spending cap, which you can specify in USD. SDV will generate synthetic data until the cap has been reached. After that, additional sampling will be paused until the cap resets at the next billing cycle. You can increase or decrease your spending cap at any time. This approach gives you predictable costs along with the flexibility to scale usage up or down as needed.

​

Billing

Subscriptions are month-to-month and can be canceled at any time. All invoices are sent to the billing email address associated with your account.

​

What information is shared with DataCebo?

SDV Enterprise is deployed entirely within your environment. After you purchase a subscription, you’ll download the software package from DataCebo and install it using your license key. All data processing, model training, and synthetic data generation occur within your infrastructure. Your data is never transmitted to DataCebo servers. To support usage-based billing, SDV Enterprise communicates with DataCebo during sampling. Before synthetic data is generated, the software estimates the number of sampled cells that will be created and sends a usage request to DataCebo. If sufficient credits remain within your configured monthly spending cap, the request is approved and sampling proceeds. After sampling completes, SDV Enterprise sends a usage confirmation containing the number of sampled cells generated. This information is used solely to update your remaining credit balance for the current billing cycle.