Pricing

Priced like infrastructure.
Not like seats.

Each pillar is priced on the value driver native to it — governed estate for Ritvam, context scope for Sutra, inference consumption for MIRA. Deployment mode is a cost multiplier, never a change of metric. No self-serve checkout — talk to us for a quote.

How we think about pricing

Platform-wide principles.

Five rules that hold across every pillar and every deployment mode.

One metric per pillar

Each pillar bills on one metric, applied identically across every deployment mode — nothing is repriced by hosting choice.

Deployment is a multiplier

SaaS is the base. Dedicated and self-hosted deployments apply a multiplier for genuinely higher cost-to-serve — never a change of metric.

No per-user pricing

Governance value scales with the AI estate under control, not headcount.

No per-call metering on control infrastructure

Ritvam and Sutra are never metered per call — compliance buyers need a predictable bill.

Predictable annual contracts

Annual minimum, with 2–3 year terms preferred and locked rates — matched to BFSI procurement cycles.

The model

One anchor. Two attach modules.

Ritvam is mandatory. Sutra and MIRA are natural extensions on top of it — not separate products competing for attention.

Ritvam

Ritvam

Anchor · Required

Billing metric

Governed AI Workloads

Billed per governed workload — every agent, application, or model endpoint registered in Ritvam and subject to its policy enforcement, audit logging, and access control.

  • Priced per workload, at the governance tier that workload requires
  • Deployment mode (SaaS / dedicated VPC / self-hosted) applies as a cost multiplier only
  • Required for every deployment — nothing runs outside it
Sutra

Sutra

Attach module

Billing metric

Context Scope Bands

An annual subscription sized in bands defined by context scope — connected data sources, indexed data volume, and entity-graph size.

  • Bands are generous by design — sized so a normal deployment doesn't hit an overage in year one
  • No metered overage — moving up a band happens at renewal, never as a surprise charge
  • Priced as an attach module, positioned below your Ritvam spend
MIRA

MIRA

Attach module · Early access

Billing metric

Committed Consumption

Committed monthly or annual inference spend at a discounted rate, with metered overage above the commitment at list rate.

  • The only pillar priced on consumption — its cost base (GPU capacity) is itself consumption-shaped
  • Deeper commitment tiers earn larger discounts against list rate
  • Behind early access until neocloud partnership terms are finalised — join the waitlist

Committing to all three pillars upfront carries a bundle discount.

Ask about bundling
Ritvam governance tiers

Standard, Elevated, Restricted.

A workload is priced at the tier it requires — defined by the risk class of what it's allowed to do, not by how big your organisation is.

Standard

Capabilities

Policy enforcement, immutable audit trail, access control, spend governance.

Typical BFSI workloads

Internal assistants, document summarisation, knowledge search.

Elevated

Capabilities

Everything in Standard, plus PII and fairness rule packs and jurisdiction-specific compliance configurations.

Typical BFSI workloads

Workloads touching customer data: KYC document processing, customer-facing chat, HR screening.

Restricted

Capabilities

Everything in Elevated, plus propose-only execution gating, human approval workflows, and IAM-scoped credential separation.

Typical BFSI workloads

Workloads that can act: transaction-adjacent agents, IT administration, finance operations.

Commercial rules

What's consistent on every contract.

The same four rules apply across every deal, regardless of pillar or deployment mode.

Bundle discount

A discount applies when you commit to Ritvam, Sutra, and MIRA together upfront.

Implementation fee

A separate one-time fee covers onboarding and integration on every deal — never folded into the subscription.

Paid pilots

Pilot engagements are paid, at a reduced fee credited against your first year — not free trials.

INR billing

Indian BFSI customers are billed in INR.

On-premises

GPU Monk

From single GPUs to full enterprise racks — GPUMonk supplies homelabs, AI developers, and IT departments with factory-new hardware and official manufacturer warranty. No compromises.

Talk to us about pricing.

We'll map Ritvam's governed-workload pricing, Sutra's context bands, or MIRA's committed consumption to your environment and give you a quote.

Request a Quote

Frequently Asked Questions

Everything you need to know about the THQ.Digital platform.