Pricing

AI governance that follows the project, not the org chart.

Unlimited developers on every plan, including the trial. Unlimited projects on every paid plan. You pay for the size of the AI estate AIMS governs, never per head and never per project.

Unlimited developersUnlimited projects on paid plansOIDC sign in on every plan
Show prices in
01 · Plans

Three plans. Developers unlimited on all of them.

Every competing product governs at organisation or user level and bills you per seat. AIMS governs at project level and does not count heads.

Early access
£0
One project, by request
Projects
1
Developers
Unlimited
Governed spend
£1,000 total allowance
Deployment
Hosted
Sign in
OIDC
  • Unlimited developers on the project
  • Full attribution: cost per feature, story and engagement
  • Board attribution across GitHub, Jira and Azure DevOps
  • OIDC sign in
  • Control: budgets, enforcement, model assignment
  • Acceleration: templates, skills and actions

Requests keep working when the allowance runs out. Attribution pauses until you upgrade.

Request access Access provisioned within one working day
Enterprise
From £3,500 per month
Fixed fee, quoted on scope
Projects
Unlimited
Developers
Unlimited
Governed spend
Unmetered
Deployment
Self host, bring your own gateway, own storage
Sign in
OIDC, SAML, SCIM
  • Everything in Practice
  • Self hosting, bring your own gateway, customer owned storage
  • SAML federation and SCIM provisioning
  • Prompt policy in blocking mode
  • Tenant isolation and custom DPA
  • Azure Marketplace private offer available
Talk to us about Enterprise

Prices exclude VAT. Billed in GBP; USD and EUR are indicative. Pay annually and get one month free on any paid plan.

02 · What you are buying

Control, attribution, acceleration.

Three capability groups. Attribution is on every plan because a trial without it proves nothing. Control and acceleration arrive with Practice.

Control

Decide what a project may do before the spend happens.

  • Per project model assignment
  • Per project budgets with enforcement
  • Request blocking when a backlog feature reaches its budget
  • Prompt policy checked against the project brief
  • Scoped short lived key brokering
  • Client owned provider accounts
Practice and Enterprise

Attribution

Know what every shipped feature cost, and who to bill for it.

  • Cost per shipped feature, per story, per engagement, per client
  • Board attribution across GitHub, Jira and Azure DevOps
  • Chargeback export
  • Seat versus API cost comparison
Every plan, including Early access

Acceleration

A governed project starts governed.

  • Repo templates with governance wired in
  • Skills and actions for the coding tools your teams already use
Practice and Enterprise
03 · The scale meter

Practice is one plan. It scales by governed AI cost.

Governed AI cost is the total monthly AI spend AIMS observes across your estate: allocated seat cost, subscription overage, and direct API and Foundry consumption.

  1. Seat cost is included in the meter, so the figure reflects your whole estate however each vendor chooses to bill you.
  2. This is a measure of estate size, not a share of savings. AIMS takes no percentage of your AI spend and adds no markup to your provider bills.
  3. Bands are assessed on a trailing three month average and move in both directions. When an engagement ends, your bill goes down.
Practice price by governed AI cost per month
Governed AI cost per monthPractice price
Up to £10,000£650
Up to £35,000£1,250
Up to £100,000£2,400
Above £100,000Quoted

What would Practice cost us?

Practice price
£1,250 / month
As a share of governed AI cost
5.0%
Per developer per month
Enter a developer count

Compare the per developer figure with the price of a single coding assistant seat.

04 · Early access

One project, £1,000 of governed spend, access by request.

Early access is £0, and it is invite only. We want a conversation with every early user, so requests are reviewed by a person and provisioned within one working day.

  • What you get. One project, unlimited developers on it, full attribution, OIDC sign in, and a total governed spend allowance of £1,000. No control features, no acceleration features.
  • The allowance is a total, not a monthly cap. An active team gets a fast evaluation. A slower one is not punished by a calendar deadline. Remaining allowance is shown in the product as a figure and a percentage.
  • When it runs out, requests keep working. AIMS sits in the request path, so we never block inference on a trial. New spend stops being ingested and the dashboard locks behind an upgrade prompt showing what you have already accumulated. You lose the record, not your working day.
  • Access is granted in batches. While in early access we prioritise firms running multiple concurrent client engagements, because that is where project level governance earns its keep.

Request access

Seven questions, one screen. Fields marked * are required.

Which AI coding tools are in use *

We provision access within one working day. No phone number, no job title, nothing that does not help us set you up.

05 · Enterprise

A fixed fee for running it your way.

Enterprise is quoted on scope: number of governed tenancies, deployment model and support tier. There is no usage meter, because when you self host or bring your own gateway we cannot observe the spend, and we will not pretend to.

Bring your own gateway

AIMS attaches to whatever you already run: Azure API Management, self hosted LiteLLM, Kong, Envoy, Cloudflare AI Gateway, or any OpenAI compatible or Anthropic Messages compatible proxy. Your gateway is already an approved part of your control plane. AIMS governs through it rather than replacing it.

Self hosting and your own storage

Run the control plane in your tenancy with prompts, attribution and audit records in storage you own. Tenant isolation and a custom DPA are part of the contract, not an add on.

SAML and SCIM

OIDC sign in is on every plan. Enterprise adds SAML federation and SCIM provisioning so joiners and leavers are handled by your directory, not by us.

Blocking mode prompt policy

Practice runs prompt policy in advisory mode because a false positive in the inference path costs developer trust permanently. Enterprise customers with a platform team that owns the policy can switch it to blocking.

Prompt insights, aggregate by default

Individual level views are off by default on every plan. On Enterprise they can be enabled behind an explicit acknowledgement, and any cohort under five people is suppressed. Your developers can adopt AIMS without wondering who is reading over their shoulder.

Procurement

Available as an Azure Marketplace private offer, so the spend can sit against an existing commitment. Direct invoicing is fine too.

Founding customers

50% off list for the first eight.

The first eight customers get 50% off list for twelve months, locked at that rate for a further twelve at renewal. In exchange we ask for a named reference, a published case study with real numbers, and a monthly feedback session for the first six months.

Practice list£650per month
Founding rate£325per month, first 24 months

Closes at eight customers or 31 December 2026, whichever comes first. The date holds.

Paid pilot

£1,500 for eight weeks across two projects

Our team does the setup. Fully creditable against your first annual contract.

Ask about a pilot
Where the money goes

One subscription. Your provider bills stay yours.

AIMS adds no markup to your model providers and takes no percentage of your AI spend. Token costs are billed by the provider to you or to your client, in their tenancy, on their books.

AIMS subscriptionGovernance and attributionControl · attribution · acceleration
Your provider billToken spend, unmarkedOpenAI · Anthropic · Foundry · client owned
Decision ready recordEvidence you can invoice againstWhat was built · what it cost · who pays
06 · FAQs

Pricing, without the small print

What does AIMS charge for?

AIMS charges a subscription for governance and attribution. Your model provider spend is billed to you, or to your client, directly by the provider. AIMS adds no markup and takes no percentage of it.

Why is Practice priced on governed AI cost rather than projects or seats?

Because both alternatives punish the behaviour we want. Per project pricing gives you a reason to merge client engagements, which corrupts the attribution data. Per seat pricing gives you a reason to leave developers outside the governed path, and every ungoverned developer is a gap in your record. Governed AI cost measures the size of the estate AIMS is looking after, and nothing else.

What counts as governed AI cost?

The total monthly AI spend AIMS observes across your estate: allocated seat cost, subscription overage, and direct API and Foundry consumption. Seat cost is included so the figure reflects your whole estate however each vendor chooses to bill you.

How do the bands move?

Bands are assessed on a trailing three month average and move in both directions. When an engagement ends and your governed spend falls, your bill goes down with it. There is no renegotiation and no surcharge for crossing a band mid month.

Why is prompt policy advisory on Practice?

Prompt policy sits in the inference path. A false positive that blocks a legitimate request costs developer trust permanently, so customers run it in advisory mode until they trust it. Blocking mode is available on Enterprise, where a platform team owns the policy.

Can our developers see each other's prompts?

No. Prompt insights are aggregate only on every plan. Individual level views are off by default, available on Enterprise only, behind an explicit acknowledgement, and any cohort under five people is suppressed.

What happens when the Early access allowance runs out?

The allowance is £1,000 of governed spend in total, not per month. When it is used up, requests keep flowing so nobody's working day stops. New spend stops being ingested and the dashboard locks behind an upgrade prompt showing what you have already accumulated.

Do you offer annual billing?

Yes. Pay annually and get one month free on any paid plan.

Which currencies do you bill in?

Billing is in GBP. The USD and EUR figures on this page are indicative equivalents so you can read the page in your own currency; your invoice will be in pounds unless we agree otherwise on an Enterprise contract.

Ready when you are

Govern every project. Count no heads.

Request access and we will set you up within one working day.

Request access