Usage views answer two questions: how much capacity you are consuming, and where model spend is allocated.
You can find them under Settings → Billing → Usage and Detailed reports.
Capacity usage
It shows slot utilisation over time — how much of the capacity you pay for is actually in use.
View it alongside queue depth. High utilisation with a long queue means you are under-provisioned. Low utilisation means you are paying for idle slots. High utilisation with an empty queue is the ideal scenario: capacity matching demand.
On plans that meter agent-hours, they are counted only while an agent is actively working. Queue time is free.
Model spend
It is broken down so you can allocate it rather than simply observing it. The key dimensions to examine are:
- By project — which codebase is using the budget
- By agent — almost always the developer agent, given how frequently it runs
- Over time — the overall pattern matters more than any single day
How to interpret this data usefully
The useful ratio is spend compared to delivered work, not spend in isolation. A month with higher spend and a proportional increase in merged pull requests means everything is working as intended.
When spend rises without a corresponding increase in delivery, the most likely causes, in rough order, are:
- Rework. Work that runs twice costs twice. Rising rework usually stems from vague specifications — see Specifications.
- Profile drift. Routine tasks being processed under a more capable profile than required increases costs at every stage. See AI profiles.
- Task sprawl. Broad tasks require more exploration to determine their own scope.
Exporting
Detailed reports can be exported for your own analysis or to support internal chargeback. If you need automated exports rather than manual ones, see Reports.
Billing history
Invoices and payment history are located under Billing history, with payment methods managed via the billing portal. See Billing history.