Pricing
License the engines.
Power your agents.
ARDA — the Universal Discovery Engine — and MatterSpace — the Universal Generation Engine — are licensed under custom contracts scoped to your agents, deployment model, and compute requirements.
What's included
Every contract includes the full platform
No feature gates. No tier limitations on discovery modes. Your contract defines compute quotas, concurrency, deployment model, and support level.
All four discovery modes
Symbolic, Neural, Neuro-Symbolic, and Causal (CDE)
All Truth Dial levels
Explore, Validate, and Publish governance tiers
Full API surface
REST API, Python SDK, MCP, CLI, webhooks, and real-time streams
BYO Scientist
Orchestrate ARDA with your own AI agents via structured sessions
BYO LLM Provider
Bring your own OpenAI-compatible endpoint for agent cognition
Flexible deployment
Cloud-hosted, self-hosted, or air-gapped installations
Deployment options
Cloud, self-hosted, or air-gapped
The same engines are available as managed cloud software or as software you run on your own infrastructure. Your agents use identical APIs, governance models, and evidence ledger formats — integration code is portable between deployments.
You license the engine — the pipeline, the algorithms, the governance stack. Models trained on your data are yours. Discoveries made with the engine are yours.
Cloud-Hosted
Vareon provisions and manages the infrastructure. Automatic hardware selection, scaling, and monitoring. Fastest path to connecting your agents to discovery and generation.
- Managed infrastructure
- Automatic GPU selection
- Usage-based compute billing
- No hardware to provision
Self-Hosted
Deploy the engines on your own infrastructure. Your agents connect locally, you supply the compute, and data never leaves your environment.
- Full data residency control
- License fee — you supply compute
- Docker-based deployment
- Portable API and integration code
Air-Gapped
Fully isolated deployment with offline license activation. No outbound network required. Your agents operate the engines inside classified environments and regulated industries.
- Zero outbound connectivity
- Offline license activation
- Classified-environment ready
- Compliance documentation included
Value
What stagnation costs.
Research programs that rely on manual analysis to find governing equations and causal structures face a compounding cost: analyst time, delayed results, and missed relationships that sit undiscovered in data that has already been collected.
ARDA shifts that bottleneck from human bandwidth to compute. A governed discovery campaign that would take a team of analysts weeks or months can run in hours, with every claim hashed, controlled, and reproducible.
For enterprise customers, governed outputs satisfy regulatory traceability, reproducible results survive peer review, and the discovery workflow scales with data volume without proportional headcount growth.
FAQ
Common questions
How is ARDA priced?
ARDA (Universal Discovery Engine) and MatterSpace (Universal Generation Engine) are sold under custom contracts tailored to your organization's scale, agent workloads, deployment requirements, and compliance needs. Contact our sales team to discuss your requirements.
Can I evaluate ARDA before committing?
Yes. We offer guided evaluation programs where you can run ARDA on your own data under a scoped engagement. Contact sales to discuss evaluation options.
What deployment options are available?
ARDA can be deployed cloud-hosted (managed by Vareon), self-hosted (on your infrastructure with a license), or air-gapped (fully isolated with offline activation). All options use identical APIs — your integration code is portable between them.
How is compute billed?
On cloud-hosted deployments, compute is billed per GPU-second consumed. On self-hosted deployments, you supply your own hardware — Vareon charges a license fee only. Compute costs and quotas are defined in your contract.
What about data residency?
Vareon has no proprietary data and never will. You license the engine — the pipeline, the algorithms, the governance stack. Models trained on your data are yours. Discoveries made with the engine are yours. Self-hosted and air-gapped deployments ensure data never leaves your infrastructure. Cloud-hosted deployments can be configured for specific regions.
Is there a self-serve tier?
Not at this time. ARDA and MatterSpace are universal engines built from first principles for your agents — significant research and engineering inventions sold under custom contracts with dedicated support. This ensures every deployment is properly scoped, configured, and supported.
† Causal Dynamics Engine (CDE) and MatterSpace are patent pending in the United States and other countries. Vareon, Inc.
Scope your deployment.
Contact our sales team to scope the right deployment, compute, and support for your organization.
Talk to Sales