Pricing

Per-agent, four tiers, no surprises

Predictable annual pricing across four tiers, scaled by the number of agents you protect. No per-call metering, no overage charges, no soft caps. Apache 2.0 source remains freely buildable from GitHub for inspection and evaluation; production deployment requires a paid tier.

Starter

First production deployment. Single team running one or a few high-value agent workflows. The fastest path from pilot to paid.

Contact
1–25 agents · annual
  • Hard cap: 25 distinct agents (rolling 90 days)
  • All product features included
  • Commercial support, 1 business day response
  • 90-day paid pilot available
  • Annual term, hot-reload upgrades

Growth

Two teams, or one team scaling. The natural second-year motion; Starter is proven, more workflows are coming online, and the upgrade cliff has to be gentle.

Contact
26–100 agents · annual
  • Hard cap: 100 distinct agents (rolling 90 days)
  • All product features included
  • Commercial support, 1 business day response
  • Quarterly check-in
  • Annual term, hot-reload upgrades

Enterprise

Large enterprise, complex compliance posture, named-account relationship. Dedicated engineering and services on top of the product, not a feature unlock, a service layer.

Contact
251+ agents · annual
  • No agent cap (scoped per agreement)
  • All product features included
  • Dedicated support engineer
  • Custom Rego authoring
  • On-prem code review
  • Contractual SLA, roadmap input
Feature parity

Same product across every tier

Starter through Enterprise ship the same gateway, the same console, and the same set of features. Tiers differ by agent capacity and the support commitment, not by feature gates. The full capability inventory lives on the product page; what each tier's support level includes lives on the support page.

Pricing principles

How we think about pricing

Five principles that shape every quote. Anchored in what procurement, security, and engineering actually need.

01

Per-agent, not per-call

We price by the number of agents you protect, defined as distinct (tenant, subject) pairs the gateway has minted tokens for, over a rolling 90-day window. Per-call metering creates the wrong incentive on a security product ("we're being charged more for using you, so we'll use you less") and procurement teams reject it.

02

Hard caps, not metered overage

When your distinct-agent count hits a tier's ceiling, the gateway returns 402 Payment Required on new mints. Existing tokens are unaffected. Tier upgrades are renewal conversations, not surprise invoices. Upgrading is a hot-reload of the license file with zero downtime.

03

Same product across all tiers

SSO, SCIM, JIT elevation, audit verify, AI-assisted Rego, SIEM forwarders, multi-tenant scoping: all in every tier. Enterprise adds dedicated support, custom Rego authoring, on-prem code review, and contractual SLAs as services, not capability gates.

04

Apache 2.0 source, paid production

The Apache 2.0 gateway, extractor, SDKs, and helm chart remain freely buildable from GitHub. Clone, inspect, build, run for evaluation. Production deployment requires a paid tier: a contractual constraint plus a license-key technical constraint.

05

Pricing transparency

"Contact" sits in the tiles instead of a number because every first deal involves a small negotiation around pilot scope, agent count, and support commitments. We'll publish public price ranges once the first three commercial agreements anchor them.

Frequently asked

Why no public price?

Every first deal involves a small negotiation around pilot scope, agent count, integration footprint, and support commitments. A flat public number would either anchor low and undercut larger deployments, or anchor high and lose smaller teams. We publish tier structure publicly and quote within one business day on request. Once the first three commercial agreements anchor the ranges, we'll publish them.

Can I try IntentGate before buying?

Yes. The Apache 2.0 gateway, extractor, SDKs, and helm chart are freely buildable from GitHub for evaluation. Spin it up on a laptop or in CI, exercise the five checks, inspect the audit chain. Production deployment requires a paid tier (a contractual constraint plus a license-key technical constraint), but evaluation is unrestricted. Paid 90-day pilots are available on the Starter tier.

What if my agent count grows mid-term?

When distinct-agent count hits the tier ceiling, the gateway returns 402 Payment Required on new mints. Existing tokens are unaffected. Upgrading is a hot-reload of the license file with zero downtime, prorated to the renewal date. No surprise overage invoices.

What's the difference between Pro and Enterprise?

Same product. Pro is the standard commercial tier with business-day support, quarterly business review, and self-service onboarding. Enterprise adds a dedicated support engineer, custom Rego authoring, on-prem code review for regulated deployments, a contractual SLA, roadmap input, and procurement-friendly terms. None of the differences are capability gates; they are service commitments.

Do you sell through partners or resellers?

Yes, on a controlled basis. Authorised partners receive wholesale pricing and may discount up to the floor set per tier; they cannot price below the floor. This protects the brand from race-to-the-bottom commodification and prevents channel conflict between partners working the same deal. Partner enquiries: contact us.

What does the license key do, technically?

The license is a signed JWT containing tier, agent cap, expiry, and customer ID. The gateway validates the signature on startup with a baked-in public key, then enforces the agent cap at mint time. Air-gapped deployments work without registry connectivity (the JWT is self-contained); cloud deployments additionally support online activation against license.intentgate.app. License files do not phone home and contain no telemetry.

What if we deploy in multiple regions or environments?

A single Pro or Enterprise license covers multiple deployment instances under one customer, provided the aggregate distinct agent count stays under the tier ceiling. Multi-region active- active and replicated audit chains are negotiated under Enterprise terms.

How does pricing scale past Enterprise?

Past 500 agents, pricing transitions from per-agent to negotiated annual under custom commercial terms. Typical structure: a base platform fee plus a per-agent component that scales down with volume. Send us your deployment shape and we'll quote.

Need something we haven't listed?

Pro services (policy authoring, SIEM wiring, custom Rego), volume discounts past 500 agents, on-prem support, paid pilot terms; we'll quote.

Talk to us