PROVERON / PRICING

Choose the right evaluation scope.

Start with repeatable tests on generated business records. Larger pilots and research work are scoped with your team before you commit.

Discuss your evaluation PILOT ACCESS BY REQUEST
CLEAR SCOPE BEFORE PURCHASEGenerated test data

Core state checks are deterministic. Billing, data scope and availability are confirmed in the written offer.

No client records in the published benchmarkNo LLM judge for state-based grading
01 / PACKAGESProposed pricing · availability confirmed by request

Plans for product teams, operators and labs.

These are proposed commercial packages, not a self-serve checkout. Your quote confirms features, scope and billing terms.

Flight Simulator Pro billing
FOR CFOs, COOs & ENTERPRISE OPERATIONS

Enterprise Assessment

$8,500 / 30-day pilot

Fixed proposed pilot price; scope agreed in advance.

Assess candidate agents against a defined business workflow.

  • A scoped evaluation of your chosen workflow and agents
  • Read-only shadow assessment only where the workflow and data allow it
  • Review of duplicate invoices, altered supplier details and approval exceptions
  • Agreed data handling and retention before any customer material is shared
  • Findings report for your operations and risk stakeholders
Discuss an enterprise pilot
FOR FRONTIER AI LABS & RESEARCH TEAMS

Research & RL Environments

From $25,000

Scoped to environment, data and integration requirements.

Build a verifiable evaluation or training engagement around a defined research question.

  • Research environment and evaluation scope agreed with your team
  • Deterministic grading for supported state-based tasks
  • A–E decision trajectories with consent and provenance fields where available
  • Export and adapter requirements confirmed in the statement of work
Contact the research team

No trial or subscription starts from this page. We confirm current availability, billing, seat count and any customer-data handling in writing first.

02 / COMPARECurrent and proposed scope

Compare the engagement types.

IncludedFlight Simulator ProEnterprise AssessmentResearch & RL
EnvironmentsPublished finance test; other scopes by requestWorkflow agreed for the pilotDefined in research scope
Test scaleUp to 50 seeds when supported by the selected runAgreed sample and evaluation planAgreed world and episode count
Shortcut checksSupported release and run checksIncluded when relevant to the workflowDefined with the evaluation
SupportSetup and access by requestProject support during the pilotResearch and integration support by scope
Trajectory exportsAvailable fields depend on the runAgreed in the pilot planA–E fields and provenance, where available
03 / QUESTIONSBefore you purchase

What to know about pricing and data.

Do published tests use client records?

The published finance benchmark uses generated records and documents. Any separate customer assessment has its own agreed data scope and handling terms. Do not send confidential records through the public contact form.

Is grading done by an LLM judge?

State-based task outcomes are checked deterministically against the generated world. This does not mean every kind of review or written deliverable has zero model or human review cost.

Can I pay by corporate card?

Payment method depends on the engagement and is stated in the written quote or order. Contact us to confirm before purchase.

How do cancellation and renewal work?

Subscription billing is not self-serve on this page. Any renewal, cancellation window or refund terms will be stated in the order before payment.

Is the 14-day trial available?

Trial access is not currently activated through the website. Ask us about pilot access and we will confirm what is available for your workflow.