Tonic pricing: what’s included in Structural vs Textual vs Fabricate, and how do plans differ?

Most teams hit the same wall: they know they need production-like data to move faster, but they don’t want a mystery bill or a black-box SKU matrix just to get started. Tonic’s pricing is designed around the actual jobs you’re trying to do—hydrate dev and test, unblock AI workflows, keep privacy continuous—not around how many buzzwords you can squeeze into a contract.

Below is a breakdown of what’s included in Tonic Structural, Tonic Textual, and Tonic Fabricate, how plans differ, and how that translates into real usage and cost.

Quick Answer: Tonic pricing is product-specific: Structural is priced by plan and the volume of structured source data you connect, Fabricate is usage-based for AI-driven synthetic generation, and Textual is focused on unstructured data pipelines ahead of RAG and training. Plans differ by scale, deployment options, and advanced governance features, not by gating core utility.

The Quick Overview

What It Is: Tonic is a synthetic data and de-identification platform covering structured databases (Structural), unstructured content (Textual), and agentic AI generation (Fabricate).
Who It Is For: Engineering, data, and AI teams that need production-like data in lower environments, labs, and sandboxes—without copying live PII/PHI everywhere.
Core Problem Solved: You stop cloning raw production data into dev and test, while still keeping referential integrity, semantics, and statistical properties intact so your applications and models behave like they do in prod.

How It Works

Tonic’s product suite maps cleanly to your data surface:

Tonic Structural connects directly to your structured and semi-structured sources (databases, warehouses, lakes), then de-identifies, subsets, and/or synthesizes them into safe, high-fidelity test datasets.
Tonic Textual ingests unstructured artifacts—documents, emails, PDFs, tickets—and runs NER-powered pipelines to detect, redact, tokenize, and optionally synthesize sensitive content.
Tonic Fabricate lets you describe the data you want via a Data Agent and generates relational synthetic datasets, mock APIs, and unstructured artifacts on demand.

Connect & Inventory (Structural / Textual / Fabricate):
- Structural connects to your DBs/warehouses, scans schemas, and maps sensitivity.
- Textual connects to storage or content systems to index unstructured files.
- Fabricate connects at the consumption layer: you specify schemas, formats, or target systems for synthetic output.
Transform & Generate:
- Structural applies deterministic masking, format-preserving encryption, synthesis, and subsetting with referential integrity.
- Textual runs NER pipelines, redaction, reversible tokenization, and synthetic replacement of entities.
- Fabricate orchestrates agentic AI to generate fully coherent datasets and artifacts that match your described domain.
Deliver & Integrate:
- Structural hydrates dev/staging environments or exports in familiar formats (SQL, CSV, etc.).
- Textual hands off safe, semantics-preserving documents ready for RAG ingestion, LLM training, or analytics.
- Fabricate exports synthetic data and mock interfaces into your dev, demo, and testing workflows.

Features & Benefits Breakdown

Core Feature	What It Does	Primary Benefit
Structural: High-fidelity de-identification & subsetting	Transforms production structured data while preserving referential integrity and statistical properties; subsets large sources down to right-sized test sets.	Hydrate dev and QA with production-shaped data without exposing real customer identities or overprovisioning storage.
Textual: NER-powered redaction, tokenization & synthesis	Detects entities (PII/PHI/PCI, etc.), applies redaction or reversible tokenization, and can replace entities with realistic synthetic alternatives.	Feed RAG pipelines and LLM training with rich, contextual documents that no longer contain sensitive real-world identities.
Fabricate: Data Agent for synthetic generation	Uses an agentic AI workflow to create relational databases, files, and mock APIs from your schema definitions and descriptions.	Get realistic, domain-specific test and demo data from scratch—no need for production access at all.

Ideal Use Cases

Best for Tonic Structural: Because it turns your production databases into safe, referentially intact test data.
- Hydrating dev/staging from prod without shipping PII downstream.
- Cutting an 8 PB warehouse into a 1 GB subset while preserving cross-table consistency.
- Running CI/CD pipelines that depend on realistic data to catch edge cases.
Best for Tonic Textual: Because it cleans up unstructured content without breaking semantics.
- Preparing documents for RAG ingestion so your LLM sees the real workflow context, not blacked-out PDFs.
- De-identifying support tickets, clinical notes, or case logs for analytics and model training.
- Applying reversible tokenization where teams need to relink back under strict access control.
Best for Tonic Fabricate: Because it generates everything from scratch where production data is off-limits.
- Spinning up realistic demo environments without any real customer data.
- Building synthetic corpora for model training in greenfield domains.
- Populating integration test harnesses with coherent, relational data that’s tailored to your schemas.

Limitations & Considerations

Structural scope is tied to connected source volume:
Tonic Structural’s annual pricing is based on the Plan you select and the amount of source data connected. Source data is defined as the total size on disk of data sources connected to Structural, excluding logs and indexes. As you connect more, you get volume discounts, but you still need to size for your actual connected footprint.
Product fit matters more than trying to stretch a single SKU:
Structural is not a replacement for Textual when you’re dealing with PDFs and emails; Textual is built for unstructured content and NER pipelines. Likewise, Fabricate is optimized for net-new synthetic data rather than de-identifying an existing production corpus. Choosing the wrong tool can limit your fidelity or add friction.

Pricing & Plans

Tonic’s pricing is designed to map to how you deploy and scale, not just how many seats you buy. While exact numbers can change over time and are quoted directly by sales, the structure is stable and predictable.

Tonic Structural Pricing

How it’s measured

Plan-based + source data volume:
Annual pricing for Tonic Structural is determined by:
- The plan tier you select.
- The amount of source data connected (total size on disk, excluding logs and indexes).
Volume discounts:
The incremental price per GB decreases as you connect more data, so teams consolidating multiple sources into Tonic get better unit economics.
What counts as source data:
- All structured/semi-structured data sources you connect to Structural.
- Tables in passthrough or truncated mode are not counted toward generator-based table metrics, but the underlying storage still informs your footprint.
- Logs and indexes are explicitly excluded from the source data definition.

Plan highlights (conceptual)

Pay-As-You-Go (Cloud Structural):
Rapid access to Tonic Structural in the cloud for simpler use cases or smaller teams:
- Ideal if you want to connect a limited dataset and start generating safe test data quickly.
- Metered usage, so you pay for what you actually run.
- Good entry point for teams piloting Tonic without a large up-front commitment.
Annual Plans (Business / Enterprise tiers):
Annual contracts for teams that rely on Structural as core infrastructure:
- Higher source data limits with built-in volume discounts.
- Advanced governance: SSO/SAML, fine-grained access control, schema change alerts, auditing.
- Deployment flexibility: Tonic Cloud or self-hosted in your VPC.
- Enterprise certifications: SOC 2 Type II, HIPAA, GDPR alignment, AWS Qualified Software.
Subsetting-Only License (Structural Subsetting + Ephemeral):
For teams that primarily need safe downsized datasets:
- Option to purchase a limited license for Subsetting only, in conjunction with Tonic Ephemeral.
- Best when your main problem is “my warehouse is enormous, but my test environment doesn’t need it all,” and you still care about referential integrity and realistic distributions.

Tonic Fabricate Pricing

Fabricate is a usage-based product centered on the Data Agent’s ability to generate fully synthetic datasets, artifacts, and mock APIs.

How it’s measured

Fabricate is generally priced by usage/credits rather than source data, because you’re not connecting a production database—you’re generating from scratch.
A simple entry plan is available:
- $29/month plan that includes $25 in usage credits; additional usage incurs metered charges.
Higher tiers scale for:
- More frequent or larger synthetic dataset generation.
- Team-based workflows, where multiple engineers and data scientists share Fabricate projects.

Where to learn more

Product details: https://www.tonic.ai/products/fabricate
Walkthrough video: https://www.youtube.com/watch?v=qAtGUNLav5k

Tonic Textual Pricing

Tonic Textual is focused on unstructured content privacy ahead of RAG ingestion and model training. Pricing follows the same philosophy: pay for the workload you’re enabling.

While the exact SKUs and numbers evolve, you can expect:

Workload-oriented pricing:
- Based on the volume of unstructured content processed (documents, messages, etc.) and required throughput.
- Scales from pilot RAG pipelines to high-volume enterprise ingestion.
Enterprise deployment options:
- Cloud or self-hosted, aligned with the same security posture as Structural (SOC 2 Type II, HIPAA readiness, AWS Qualified Software).

Plan Fit by Scenario

Structural – Pay-As-You-Go:
Best for individual teams or early pilots needing quick access to cloud Structural without a long procurement process.
Structural – Annual (Business/Enterprise):
Best for mid-size to large organizations connecting multiple production systems and needing integrated governance, CI/CD integration, and predictable annual budgets.
Structural – Subsetting-Only + Ephemeral:
Best for teams with huge warehouses who need safe, smaller, but still referentially intact datasets for dev, QA, and on-demand ephemeral environments.
Fabricate – $29/month + metered usage:
Best for builders and smaller teams that want to prototype with fully synthetic data, or create demo datasets, without touching production.
Fabricate – Higher usage tiers:
Best for product teams and data science groups that rely on Fabricate to continuously generate synthetic data for testing, demos, or training.
Textual – RAG/AI-focused tiers:
Best for data and AI teams standardizing a privacy pipeline across documents, tickets, emails, and records ahead of RAG ingestion or LLM training.

Frequently Asked Questions

How does Tonic Structural pricing actually scale as I connect more data?

Short Answer: Structural is priced by plan and connected source data volume, with built-in volume discounts so your marginal cost per GB drops as you add more sources.

Details:
Structural treats source data as the “size on disk” of connected databases, warehouses, or other structured sources, with logs and indexes excluded. When you move from a small pilot (one or two databases) to a consolidated footprint (multiple systems or a large warehouse), the per-GB price decreases. This is intentional: the more you standardize your test data pipeline on Structural, the better your unit economics.

This model encourages teams to stop running ad-hoc DIY pipelines on the side and centralize de-identification and subsetting into one governed tool, instead of having ten slightly broken scripts each cloning sensitive data differently.

Can I use Tonic Subsetting on its own, without the full Structural feature set?

Short Answer: Yes. You can purchase a limited license for Subsetting only, used together with Tonic Ephemeral.

Details:
If your immediate pain is, “Our warehouse is enormous, and every full refresh crushes our staging cluster,” you don’t need to turn on every Structural feature on day one. Tonic supports an option to license Subsetting only, paired with Tonic Ephemeral to spin up safe, smaller, but production-shaped datasets on demand.

You still get referential integrity and realistic distributions, but you’re not committing to the broader set of transformations and workflows until you’re ready. This is a common entry point for teams with extremely large environments—think multi-petabyte Snowflake or on-prem warehouses—who want a quick win on cost and speed first.

Summary

Tonic’s pricing is aligned to what you’re actually trying to ship:

Structural gives you high-fidelity, referentially intact test data from production sources, priced by plan and connected source volume with volume discounts as you scale.
Textual turns unstructured content into safe, semantics-preserving inputs for RAG and LLMs, priced around content throughput and enterprise deployment needs.
Fabricate lets you generate synthetic data from scratch using a Data Agent, with a low-friction $29/month entry point plus metered usage for heavier workloads.

Underneath the SKUs, the principle is simple: accelerate development and AI initiatives with production-like data, without leaving a trail of uncontrolled, sensitive copies across your environments.

Next Step

Get Started