Company / Pricing
Priced on governed agents. Not on how safe you are allowed to be.
Capacity scales with the ladder. The governance posture does not — fail-closed behaviour, the detector set, every enforcement point and the evidence ledger are the same on the entry paid tier as on Enterprise.
From $29 per month Agents, not seats or models Enforced entitlements, tested against this page
THE LADDER
Four tiers and a custom track.
The capacity number is concurrent governed agents. Retention is how long the evidence ledger holds governed-action records.
Free $0/ month
- Agents
- 2
- Policies
- 5
- Events / mo
- 25K
- Retention
- 7 days
The plan a new organisation starts on. Quotas enforced, no card.
Startup $29/ month
- Agents
- 10
- Policies
- 25
- Events / mo
- 200K
- Retention
- 30 days
The entry paid tier — and where the governance surfaces were widened down to.
Recommended Pro $99/ month
- Agents
- 50
- Policies
- 100
- Events / mo
- 1M
- Retention
- 90 days
Adds SIEM integration and custom dashboards.
Team $299/ month
- Agents
- 200
- Policies
- 500
- Events / mo
- 5M
- Retention
- 180 days
Identical capabilities to Pro. What changes is capacity.
Enterprise Custom
- Agents
- Unlimited
- Policies
- Unlimited
- Events / mo
- Custom
- Retention
- 365 days
SAML SSO, SCIM provisioning, federation, delegated admin, private deployment, advanced evidence export.
Managing agents for more than one customer Partner Standard adds child accounts, not a different governance posture. The plan set defines a partner tier carrying everything in Pro plus up to 10 child accounts, 100 governed agents, 2M events a month and 90-day ledger retention. It adds a partner admin dashboard, consolidated billing across those accounts, white-label policy templates, and partner API access. The fail-closed behaviour, the detector set, every enforcement point and the hash-chained ledger are the same here as on every other paid tier — the table above applies unchanged.
Described as defined, not as offered. These are the entitlements in the shipped plan set. Self-serve signup is not open for this tier any more than for the others, so the route in is the same briefing as everywhere else on this site. “White-label” here means policy templates specifically, not partner-branded console or reports.
Talk through a partner deployment →
Planned, not yet available. The billing design defines a 14-day card-required trial that converts to Startup on day 14, but it is not an entitlement in the shipped plan set and self-serve signup is not open. It is described here as designed, not offered. Nothing is entitled before payment confirms — a plan sits inactive until its checkout completes rather than granting capacity optimistically, and a failed webhook never defaults to active.
CAPABILITY PARITY
Core governance starts at Startup. Capacity and support scale from there.
Trial is a bounded evaluation path. Startup, Pro and Team share one verified governance foundation — the commercial fences were deliberately flattened toward the lower tiers, under a rule that they may only ever widen or hold, never narrow on an existing subscriber. Enterprise is tailored deployment and terms, not a way to buy basic safety.
Included everywhere
Why it is not a tier lever
The pre-production gate
Assurance runs before an agent ships, not as an add-on once it has. Six domains, bars registered before the run, and a CI/CD exit code that can block a release. Gating the gate behind a tier would contradict the sentence at the top of this page.
Fail-closed enforcement
Any transport error, timeout, non-2xx or unparseable response resolves to deny. A safer failure mode is not an upgrade.
The full detector set
Secrets, sensitive data, regulated identifiers, injection, encoded payloads, risky tool intent and unsafe destinations run identically on every paid tier.
Every enforcement point
Gateway, transparent MCP proxy, forward-auth and SDK. What scales is how many agents you govern through them, not where you may insert.
The hash-chained ledger
Evidence integrity is structural. A cheaper plan does not get a weaker audit trail.
opa_policies — keep your own engine
Bring the OPA/Rego you already run. Deliberately widened down to the entry paid tier rather than held as an upsell.
Cloud providers and MCP catalog
Connection, discovery and the registry, from Startup upward.
Compliance reports and webhooks
From Startup upward.
MFA and custom roles
Access control is not a premium feature.
Governing an agent safely is the product. Charging more for the safe failure mode, the detectors or the audit chain would make the cheapest plan the most dangerous one to run — which is not a ladder we are willing to build.
WHAT DOES DIFFER
Capacity, integration depth and support.
Dimension
Where it changes
Capacity
Governed agents and evidence retention, per the table above.
Pro and above
SIEM integration, custom dashboards.
Team
Capacity only — Team grants no entitlement Pro does not already have.
Enterprise
SAML SSO, SCIM provisioning, SSO enforcement, org-wide MFA enforcement, multi-cluster federation, delegated admin, private deployment, advanced evidence export.
HIGH TOUCH
Two tracks that start with a conversation.
Design partner
A scoped six-week evaluation on one consequential agent workflow, with named success criteria and founders in the room. The GovernorAI licence is waived; the infrastructure it runs on is yours and is not free, and we size it with you before you commit. Production is priced separately and only if you continue. What we ask in return is a reference — a named case study and permission to use your logo — contingent on the result, once the engagement has concluded and you are willing to stand behind it.
How the programme runs.
Enterprise
Priced per governed agent, so the bill tracks the same capacity dimension the tiers above use rather than a separate metric. Adds federation, delegated admin, private deployment and separation of duties on policy approval.