Skip to main content
The usage dashboard in Settings → Usage shows where your credits go: charts over time, per-category and per-model breakdowns, and a live event log. All values are in credits.
Usage dashboard showing the cost-over-time chart, date controls, and summary cards

​
Credits remaining

The top of the dashboard shows Credits remaining as used-versus-cap for the current month (for example 87.50 / 100), combining your plan allowance and any active top-up.

​
Usage categories

Every charge falls into one of five categories:
CategoryWhat it covers
AI BuilderAI builder generations and edits
Agent / LLMAgent nodes and in-workflow LLM calls
Third-party APIMetered APIs such as X and scraping
Compute (CPU)Sandbox CPU time (not currently charged)
Compute (GPU)Sandbox GPU time (not currently charged)

​
Charts and views

Two toggles control the main chart:
  • Bar / Pie: a stacked cost-over-time bar chart, or a donut showing each slice’s share of total spend.
  • By Resource / By Model: group spend by the five categories, or by individual model and service (for example a specific LLM, Veo, or a scraper).
The legend is sorted by spend with a total per entry; hovering an entry highlights it in the chart. Summary cards below the chart show Total Cost and per-category totals for the selected period.

​
Date ranges

Use the 7d / 30d / 90d presets, or the date picker next to them to select any custom range.

​
Recent usage events

The Recent Usage Events table lists your latest charges with timestamp, category, resource, count (for example tokens), and cost. New events appear in real time while your workflows run, so you can watch a run’s spend as it happens.
Recent usage events table showing timestamps, categories, resources, and costs

​
Top models and services

The Top Models & Services table ranks the most expensive entries in the selected period by day, with tokens and cost per row. Use it to spot which model or service drives your spend.

​
Next steps