AI usage & budgets

Most of Vibel runs on plain rules and math, checking a stock level, comparing this week to last week, that costs nothing. A handful of features call an AI model: the morning brief and weekly digest write-ups, deeper analysis on request, and the weekly AI-visibility probing that asks ChatGPT, Claude, Gemini and Perplexity your tracked questions. AI usage (/settings/ai-usage) is where you see the bill for all of it, broken down by feature.

Every brand has a daily spending cap and a monthly spending cap on AI calls, shown as two meters at the top of the page: how much you have spent today and this month, against each limit. The bar turns amber near the limit and red once you are over it.

Nothing breaks. Once the daily or monthly cap is hit, AI-powered features fall back to their plain, deterministic version automatically, the morning brief still sends with your real numbers, AI-visibility probing simply pauses until the cap resets, then picks back up on its normal schedule. You lose the AI polish, never the underlying feature.

The AI-visibility capability meters show how many tracked prompts and how many AI engines you are using against your plan's allowance, separate from the dollar caps above them. "Last 7 days" gives a quick trend, and "Split by feature" breaks this month's spend down by exactly what used it, so you can see at a glance whether it is the visibility probing, the morning brief, or something else driving the number.