licensing.tiers
You pay for what you govern
Three licensing tiers matched to real usage, operational criticality and scale. No features held back, and the infrastructure inside the price.
Prices per governed entity, per month.
Licensing plans
Standard
For bringing a first agent estate under control, with everything you need included from day one.
€47.50per governed entity / month
Start with Standard- Governed entities
- up to 25
- RAG queries per month
- ~10,000
- Tokens per month
- ~5M
- Actions per hour
- up to 100
Infrastructure included
2–4 vCPU · 8–16 GB RAM
Premium
recommendedFor an estate already in production: more volume, more speed and a lower unit price.
€41.25per governed entity / month
Talk about Premium- Governed entities
- up to 100
- RAG queries per month
- ~30,000
- Tokens per month
- ~15M
- Actions per hour
- up to 300
Infrastructure included
8–12 vCPU · 32–64 GB RAM
Enterprise
For large or critical estates, with dedicated infrastructure and bespoke terms.
Custom
Request a proposal- Governed entities
- 100+
- RAG queries per month
- 100,000+
- Tokens per month
- 50M+
- Actions per hour
- 1,000+
Dedicated infrastructure
16+ vCPU · 128 GB+ RAM
billing.unit
A single unit for everything you govern
There is no one rate per agent, another per connector and another per data volume. There is one unit — the governed entity — and the only thing that determines your plan is how many you have. Budgeting stops being a guessing exercise.
It counts everything you connect and bring under governance, not just the agents. If it is inside the perimeter and you govern it, it takes up an entity; if you disconnect it, it stops taking one up.
- A single licensing unit, with no catalogue of extras
- The price per entity falls as you grow
- Infrastructure included in Standard and Premium
What takes up an entity
- Platform users
- Your own agents and third-party agents
- Connected systems
- IoT devices
- Data sources: email, shared drives, databases, data lakes
The platform's own agents count too, and you can switch them on or off as suits you.
pricing.degressive
The price per entity falls as you grow
Premium costs 13% less per entity than Standard. That is not a discount for negotiating: it is in the rate card.
It is deliberate. A model that makes every new agent more expensive ends up pushing agents out of the registry, and an incomplete inventory is exactly the problem you came to solve. Governing more has to pay off.
plan.parity
The plans differ in capacity, not in functionality
There are no paid add-on modules and no features reserved for the superior plans. What changes between tiers is how many entities you govern and at what volume, not what you can do.
All four modules, complete
Telemetry, asset registry, governance and analytics. Nothing trimmed by plan.
The three integration plans
OpenTelemetry, REST API and MCP. Connect wherever you meet the least friction.
EU AI Act coverage
Per-agent mapping to Articles 12, 13, 14 and 15, with status derived from real data and export to CSV and PDF.
Verifiable reports
Daily, biweekly or monthly, versioned and sealed with the fingerprint of the dataset.
Quality evaluation
The five LLM-as-Judge dimensions over whichever executions you decide to evaluate.
Analytical chat
Natural-language queries over your telemetry, with a read-only SQL agent. Enabled per deployment.
tier.enterprise
When the estate is large or the process is critical
Above a hundred entities the conversations change: they stop being about limits and start being about isolation, reserved capacity and fit with your architecture.
- Dedicated infrastructure, sized with you and not shared
- Query, token and throughput volumes defined against your real load
- Fit with your deployment: the three integration plans work the same at this scale
- Bespoke agents are built with AKP, which ships with the platform connectors built in
faq.pricing
Frequently asked questions
What counts as a governed entity?
Everything you connect to the platform and bring under governance: users, your own and third-party agents, systems, IoT devices and data sources — from a mailbox or a shared drive to a database or a data lake. The platform's own agents count too, and you can switch them on or off. The advantage of a single unit is that the budget does not depend on guessing the mix right.
Why does Premium cost less per entity than Standard?
Because the unit price falls with scale, and that sits in the rate card rather than in the negotiation. If governing the thirtieth agent cost the same as the first, the incentive would be to leave it out of the inventory — and an incomplete registry is no use for governing or for demonstrating compliance.
Are there features only available on the top plan?
No. The four modules, the three integration paths, the EU AI Act mapping and the reports are on all three plans. What changes is capacity: how many entities, how many queries, how many tokens and at what rate.
Is the infrastructure really included?
Yes on Standard and Premium: the resources listed in each plan are inside the price and do not turn up as a separate cloud bill. Enterprise runs on dedicated infrastructure, sized with you.
How much work is it to start?
Register the agent, copy its key and point the exporter or make a POST. If your agents already emit OpenTelemetry there is no integration to write. What usually takes longer is not connecting, but agreeing internally on which agents exist.
What if I also need to build the agents?
That is the other half. AKP is the kernel Social Labs uses to deliver bespoke agents for large clients, and it ships with the connectors to Regentic.AI: point them at your instance to switch them on. You can contract the governance alone, the build alone, or both.
Contracting terms and usage
{{WIP: contracting terms — applicable taxes, billing frequency, minimum term, policy on exceeding consumption limits and plan-change conditions — pending confirmation}}
Let us put numbers on your case
Tell us how many agents you have, which of them are critical and what compliance is asking of you. The plan follows from that.