Skip to main content
The AI Usage dashboard lets you monitor how much of your AI plan your application is consuming. It shows monthly and daily progress against your plan’s limits, a breakdown of token usage by type, recent activity, and a daily-consumption chart so you can spot spikes before they block your users. Use this page to understand when to check the dashboard, what each metric means, and how the new daily limit interacts with your monthly budget.

​
When to use it

Open the AI Usage dashboard when you need to:
  • Check how close your application is to its monthly AI limit.
  • Investigate why an AI request was blocked with a daily-limit error.
  • Audit how much of your consumption comes from input, cached input, output, or embedding tokens.
  • Compare AI usage across past months for capacity planning.

​
How to open the dashboard

The dashboard is exposed as the AI Usage button in the Settings drawer.

​
Prerequisites

  • You must be signed in with a role that has access to the Feature Designer. The AI Usage button is hidden for roles without that permission.
  • Your application must have an AI package assigned (Included, Starter, or Pro) for limits and percentages to be displayed.
  1. Open the Settings drawer from the application toolbar.
  2. Click AI Usage. This takes you to /ai/usage for the current application.
  3. Use the month picker in the header to switch between the current month and any past month.

​
Daily vs. monthly limits

AI consumption is now enforced on two levels:
LimitHow it’s calculatedWhat happens when reached
MonthlyThe valueLimit defined by your AI plan.All AI requests are blocked until the next billing cycle or until you upgrade your plan.
DailyvalueLimit ÷ 30 (the plan’s per-day budget).AI requests are blocked for the rest of the day. They resume automatically tomorrow.
Both limits are tracked independently. Reaching the daily limit does not consume your monthly notifications (70%, 90%, 100%); the monthly counter keeps accumulating across days as usual.

​
Daily-limit error message

When an application has already consumed its per-day budget, the next AI request fails with:
Se alcanzó el límite diario de uso de IA de tu plan. Intenta de nuevo mañana. (The daily AI usage limit for your plan has been reached. Try again tomorrow.)
If you see this message but the monthly progress bar is well below 100%, the application has hit the daily ceiling. Wait until the next day, or upgrade the plan if your team consistently needs more daily headroom.

​
Reading the dashboard

The dashboard is organized into the following sections.
  • AI Plan tag — The AI package currently assigned to the application (Included, Starter, or Pro).
  • Period picker — Switch the entire dashboard to a different month. Future months are disabled.
  • Shared billing banner — Shown when the application’s AI usage is billed to a parent application. The banner names the billing application so you know where the consumption is being attributed.

​
Threshold alerts

Banners appear at the top of the dashboard when the monthly usage crosses these thresholds:
  • 70% — Warning that the application has used a large share of its monthly budget.
  • 90% — Stronger warning that the limit is close.
  • 100% — The monthly limit has been reached. A button to contact support to upgrade the plan is shown next to the alert.

​
Monthly usage card

  • Value consumed — A progress bar showing the percentage of the monthly limit used so far.
  • Total consumed — The accumulated value spent this month.
  • Monthly limit — The plan’s valueLimit.
  • Projected end-of-month — Only shown for the current month. Extrapolates the current daily average across the remaining days and highlights in red if the projection exceeds your monthly limit.

​
Today’s usage card

  • Daily progress bar — Percentage of today’s per-day budget consumed. The bar turns yellow at 70% and red at 90%.
  • Today consumed — Value spent so far today.
  • Daily limit — valueLimit ÷ 30, displayed alongside a tooltip explaining how it’s derived.

​
Token breakdown

Cumulative token counts for the selected month, split into:
  • Input tokens — Tokens sent to the model in your prompts.
  • Cached input tokens — Input tokens served from the provider’s prompt cache (typically cheaper).
  • Output tokens — Tokens the model generated.
  • Embedding tokens — Tokens used for vector embeddings.

​
Activity

  • Total requests — Number of AI requests recorded this month.
  • Last used — Timestamp of the most recent AI call.

​
Daily consumption chart

A bar chart of value consumed per day for the selected month, with a dashed reference line at the daily limit so you can see at a glance which days hit the ceiling. If there is no consumption data for the period, the chart shows an empty state instead.

​
What to do when you hit a limit

  • Daily limit reached — Wait until the next day. AI calls resume automatically. If your team needs more headroom every day, ask your administrator to assign a higher AI package.
  • Monthly limit reached — Contact support@codenull.io (or use the Upgrade plan button in the 100% alert) to move to a larger AI package, or wait until the next billing cycle.
Cached input tokens are tracked separately because they typically cost less than fresh input tokens. A high ratio of cached input tokens to input tokens is a good signal that your prompts are reusing context efficiently.