Pricing
A plan, plus what your agents actually use.
Per-seat pricing makes no sense here: two agents and forty agents are not the same product. So there is a plan price and metered usage — and the product estimates what an agent will cost before you deploy it rather than after your first invoice.
Builder
For trying it properly, not a demo.
Free
Free
Individuals, developers and anybody evaluating this before committing.
- Agent Studio and the guided builder
- Full sandbox with synthetic data and mock tools
- SDKs, API, CLI and the agent manifest
- Up to 3 agents, development environment only
- Dry runs and simulation against sample history
- Evaluations, runnable locally and in CI
- Run timelines and the full audit trail
What is metered
Nothing is metered, because nothing here reaches a production system. Sandbox runs are included and there is no card required.
Business
For agents doing real work.
€149
per month, plus usage
Companies deploying their first operational agents against live systems.
- Everything in Builder
- Production agents, with staging and production environments
- Permissions, approvals and the approval inbox
- All connected business tools and custom tools
- Identity per agent, secrets vault and network policy
- Budgets, limits and the emergency stop
- Agent directory, observability and impact reporting
- Unlimited agents and unlimited people
What is metered
Metered on model usage and runtime — what your agents actually consume. Every agent shows an estimated monthly cost before you deploy it, and its actual against that estimate afterwards. A typical first agent runs at €10 to €25 a month.
Scale
For many agents across many teams.
€449
per month, plus usage
Companies running agents across several departments, with their own policies.
- Everything in Business
- Custom policies and organisation-wide authority rules
- Multiple teams, business units and scoped environments
- Agent teams and delegation
- Advanced observability, retention and export
- Permission review workflows and separation of duties
- Priority runtime capacity
- Publishing to the marketplace
What is metered
Same metering as Business, at a lower rate, with a monthly usage allowance included. Workspace-wide cost ceilings are available so a misconfigured agent cannot produce a surprise.
Enterprise
For requirements that are yours specifically.
Custom
Talk to us
Organisations with data residency, networking or assurance requirements.
- Everything in Scale
- Data residency and dedicated capacity
- Custom network boundaries and private connectivity
- Your own model providers or hosted models
- Extended retention and legal hold
- Security review support and named contacts
- Commercial terms to match your procurement
What is metered
Committed usage at an agreed rate, with the same estimate-before-deploy behaviour. We will tell you what we think you will spend before you sign, and we would rather be roughly right than optimistic.
Prices exclude VAT. Model and runtime usage is metered and shown per agent, with an estimate before you deploy — see any agent's cost panel. No plan meters people: the number of colleagues who can review an approval is never a reason to pay more, because an approval queue only one person can reach is how approvals get missed.
Before you deploy
You see the cost on the same screen as the permissions.
Both are things you are agreeing to, so they appear together. Runs per month, model usage, runtime and a total — and afterwards, the actual against that estimate.
You do not estimate tokens
Predicting a bill should not require you to model token consumption. The platform does that arithmetic and takes the risk of having done it wrong.
A wrong estimate is our bug
If an agent costs materially more than its estimate, that is a defect in the estimate. We would rather you saw the number at the review screen and held us to it.
A ceiling stops things
Cost limits are enforced by pausing the agent, not by an email after the fact. An agent with a €120 monthly cap does not spend €121.
People are never metered
The number of colleagues who can review an approval is never a reason to pay more. An approval queue only one person can reach is how approvals get missed.
Estimated monthly cost
Executive Briefing Agent · before you deploy it
Shown before you deploy, not after your first invoice. If an agent turns out to cost more than this, that is a bug in the estimate and we would rather you saw it here.
Two real figures, for scale.
A morning briefing agent — twenty-two runs a month, reading several systems and writing one email — runs at about €11 a month. A finance agent making over a thousand decisions a month, chasing invoices and escalating exceptions, runs at about €74.
Those are the two ends of what most first agents look like. If you tell us what work you are thinking of handing over, we will estimate it before you sign anything.
Questions
About pricing.
Do I need a developer?
No. A business user describes a job in plain words and the studio builds the agent, and every capability, boundary and approval rule is shown in words before it is deployed. Developers get SDKs, an API, a CLI and manifests for agents that need real logic. Both produce the same kind of agent with the same permissions and the same audit trail — there is no weaker path.
What does it actually cost to run an agent?
A plan, plus what your agents consume in model usage and runtime. The product estimates an agent's monthly cost before you deploy it — runs, model, runtime and a total — and shows the actual against that estimate afterwards. A light agent such as a morning briefing runs at around €11 a month; a busy finance agent doing over a thousand decisions runs at around €74. If an agent costs materially more than its estimate, that is a bug in the estimate and we would rather you saw it at the review screen.
Can we try it without risk?
Three ways, in increasing order of how much they will convince you. A sandbox with synthetic customers and mock tools, where there is no route to a production system at all. A dry run that tells you what the agent would have done yesterday against your real records. And a simulation across a month of your history — how many customers it would have contacted, escalated or left alone. None of them sends anything or changes a record.
What if we want to leave?
Your agents, manifests, run history and audit trail export on every plan, throughout the term and during any wind-down. Manifests are declarative files you can keep in your own source control from the first day, so the definition of what your agents do is never only in our database.
Anything else, write to [email protected] and you will get a number rather than a discovery call.
Give AI a job.
Build an agent, choose what it can access, define what it may do, and deploy it to work safely across your business.
No code required · Full SDKs and API · Stop any agent instantly · [email protected]