AI allowance

AI features consume a token allowance that comes with your plan. User Profile, under Plan and access, shows AI tokens as how much you have used of your allowance and how much is left, or that your plan has no allowance.

When a feature is not included in your plan, the assistant shows the reason and Choose your tier. When the allowance is exhausted, the plan dialog opens with a heading about the allowance being reached.

What consumes the allowance: assistant conversations, voice mode, strategy suggestions, the mental game coach, import mapping assistance, and deep reviews. Deep reviews cost the most by a wide margin, because they query your data many times. That is another reason the cadence is monthly rather than weekly.

What does not consume it: every calculation the dashboards do. Metrics, charts, rule scoring, and Gravitas are computed locally and in your own pipeline, not by a language model. Asking the assistant for your profit factor uses tokens for the conversation; the number itself is computed the same way the card computes it.