perathos

// pricing

Priced for enterprise deployment.

Perathos is sold as an annual enterprise contract. Pricing is based on verified call volume and deployment configuration.

pilot

Proof of Concept

90-day scoped engagement. We deploy Perathos in your environment, run against your actual AI workloads, and demonstrate ROI before you commit.

  • Up to 50k verified calls/month
  • 2 industry circuits included
  • Dedicated onboarding engineer
  • Full proof bundle access
  • Weekly verification reports
Start a Pilot →
most common

enterprise

Enterprise Annual

Full deployment with SLAs, dedicated support, and custom circuits for your domain. Volume-based pricing with predictable annual contracts.

  • Unlimited verified calls
  • Custom industry circuits
  • 99.9% uptime SLA
  • Dedicated customer success
  • On-premises / private cloud option
  • Custom knowledge base integration
  • VRL audit chain export
  • SSO & enterprise security
Contact Sales →

platform

Build On Perathos

For teams building AI products and want to offer verification as a feature. API access to the full judge pipeline.

  • Full API access
  • White-label options
  • Circuit marketplace access
  • Reseller agreements available
  • Dedicated API engineer
  • Revenue share model
Talk to Us →

Common Questions

How does pricing scale with volume?

Pricing is based on monthly verified call volume in tiers. Most enterprise customers operate in the 500k–10M calls/month range. We structure annual contracts with overage protection.

Can we deploy on-premises?

Yes. Perathos can be deployed in your private cloud or on-premises environment. The verification pipeline and all AI calls stay within your network perimeter.

What happens to the data Perathos sees?

Perathos sees AI responses in transit. We do not store response content beyond the verification process. VRL Proof Bundles — the cryptographic verdicts — are retained for audit purposes. Data processing agreements available.

How long does verification add to latency?

Typical verification latency is 800ms–2s depending on the number of LLM-based verifiers active. Deterministic-only configurations (symbolic solver + schema validator) add <50ms. Configurable per deployment.

Can we bring our own knowledge base?

Yes. The knowledge graph verifier is pluggable. You connect your domain database (regulatory databases, drug formularies, legal corpora) and Perathos queries it for factual verification.