The right model for each task
AI spend control with caps per team
You pay for real usage, with caps per person and team. No more unused licences or tokens burned on tasks that never needed them.
Updated
From licences to consumption
The usual model is seats: you buy a hundred and twenty-five people use them. With centralised access you pay for what is actually consumed, and those who do not use the tool cost nothing. The saving does not come from a discount, it comes from no longer paying for air.
Caps that avoid surprises
You set a limit per person and per team, with a warning as it approaches and either a downgrade to a cheaper model or a stop when it is reached. Nobody discovers the bill at the end of the month: the bill is decided at the start of it.
The right model for each thing
Summarising an email does not need the most expensive model in the catalogue; analysing a long contract probably does. When every request goes through one place, that rule can be applied systematically instead of being left to each person's judgement, and that is usually where the biggest difference in the bill sits.
No promised number
We are not going to tell you that you will save a specific percentage: it depends on how many people use AI today, how, and on which licences. What we do is measure it during the pilot and show you the comparison against what you pay now, which is the only figure worth anything.
An example with made-up numbers
A company of 100 people buys 100 licences at €25 a month and finds that 25 use them. That is €2,500 a month for 25 people: €100 per active person. With centralised access only the consumption of those who use the tool counts, with a cap per person and per team. The numbers are illustrative: the real saving is measured in the pilot with your data.
What happens when the cap is reached
There are three options and you choose: warn and carry on, switch automatically to a cheaper model, or pause until the next period. Caps apply per person, per team and for the whole company, with a warning when the percentage you decide is reached.
Frequently asked questions
How do you control a company's AI spend?
By setting a cap per person, per team and for the whole company, and deciding what happens when it is reached: warn, switch to a cheaper model or pause. All usage goes through a single point, so the limit can really be enforced.
How much will I save with Acotia?
We do not give a percentage because it depends on how many people use AI today and with which licences. In the pilot we measure your consumption and show you the comparison with what you pay now.
Can the model each person uses be chosen?
Whoever administers can set a default model and stop each person from choosing. The panel also estimates how much would be saved by using a light model on short requests.
- Company cap€1,000 / month
- Cap per person€40 / month
- Warningat 80%
- On reaching 100%cheaper model