Usage & consumption

The consumption page shows how your project uses thinkhx over time, measured two ways: the reasonings run in a period, and the tokens those reasonings consumed. You can read either metric as a total, as a chart broken down day by day, and as a ranking of your applications and models. Use it to see where your allowance goes and spot trends.

Review your usage

1. Open the consumption page.
2. Read the total number under Reasonings for the selected period.
3. The chart shows one bar per day (or per week for long ranges). Click a bar to see that day's breakdown by status.

The legend shows three statuses: Finished (completed runs), New (runs that haven't finished) and Error (runs that ended in error).

Switch between reasonings and tokens

Your usage allowance is metered in tokens, not in reasonings, so the page counts both. Switch metric to read the same period the other way.

1. At the top of the chart, click Tokens.
2. The total becomes the tokens consumed over the period, and the chart bars become tokens per day.
3. The legend becomes Input (the tokens you sent) and Output (the tokens the model produced).
4. Click Reasonings to switch back.

The rankings below the chart follow the metric you chose, so the switch changes the whole page and not just the chart.

Note — A reasoning that ends in Error consumes no tokens, so it counts in the Reasonings total but adds nothing to the Tokens total. The two metrics answer different questions and deliberately do not share a filter — expect the token chart to be empty on a day where every run failed.

Note — The Tokens switch only appears once your project has runs metered in tokens. Until then the page shows the reasoning count alone.

Change the date range

1. Click the date range at the top of the page.
2. Pick your own start and end dates, or click the Last 30 days preset.
3. Click Apply.

The chart and totals update to the range you chose. The default is the last 30 days.

Find your heaviest apps and models

Below the chart, when there's activity in the range, two cards rank your usage:

1. Applications — your top applications for the selected metric.
2. Models — your most-used models for the selected metric.

Both cards are ranked by whichever metric is selected above, so switching to Tokens re-ranks them: your heaviest applications by tokens are not always your busiest ones by number of reasonings.

Tip — To check how much of the current allowance is left, look at the Plan usage card on your dashboard, which also shows the date it resets under Reset on. If you are running short, raise it with a higher plan, or with a one-off pack if you are already on the Business plan — see Plans & billing.

Note — Your allowance runs on a cycle anchored to the date your plan was activated, not on the calendar month, so a total read here for "this month" won't match the gauge unless you pick the same dates. See Plans, usage & limits.