Skip to main content

How AI credits work and what consumes them

One credit, one AI feature

Every use of an AI feature consumes one credit. An AI Field computed on a meeting is one credit. The other AI capabilities draw on the same pool in the same way.

There is no separate budget per feature, and no feature that is free. That makes the arithmetic simple and slightly unforgiving: your consumption is the number of AI operations you run, multiplied by your meeting volume.

Four AI Fields across two hundred meetings a month is eight hundred credits. That multiplication is the calculation most teams have not made before they turn fields on.

Where your allowance comes from

Your allowance is set by your licences. Each licence grants a monthly amount of credits, and the total is that amount multiplied by your number of users.

The credits are then pooled at workspace level rather than held individually. It is the whole team's usage that draws on the same envelope, which means one heavy user can consume what the rest of the team was relying on. There is no per-person counter to protect anyone.

A practical consequence: adding or removing licences changes your credit allowance too, not just your seat count.

When it renews

Credits renew every month on your workspace's billing date, not on the first of the month.

If you are billed on the 3rd, your credits reset on the 3rd. This matters more than it sounds, because a team that plans its analysis in calendar months and then reads a balance mid-cycle will conclude the numbers are wrong when they are simply reading a different window.

Your current balance and the exact reset date are shown in Workspace Settings, then Plans & Licenses.

Before you turn on more

Three things worth doing in order.

Count your monthly meeting volume first, then multiply by the number of AI operations you intend to run on each. That gives you your real consumption, and it is almost always higher than the intuition.

Start with three or four AI Fields that map to something you already report on, rather than everything you could extract. A field you never look at costs exactly the same as one you use daily.

Check your balance before adding fields mid-cycle rather than after. A field that runs out of credits partway through the month leaves you with an incomplete column, which is worse than not having the field at all.

Did this answer your question?