Every job you run is divided into one or more tasks. We track container usage throughout each task's lifecycle—not the total container time—to ensure a fair calculation of your job's price.
We measure actual usage statistics, such as CPU time and average RAM consumption within your container, rather than relying on the overall machine capacity. These metrics are then matched with the rates listed on our website to calculate the cost for each billable process across your job's lifecycle.
The following image illustrates what qualifies as billable process:
Unlike other services, we do not charge for idle container time or machine pre-warming. Each app you build includes guaranteed idle capacity, reducing cold starts and ensuring smooth execution. This means you only pay for the time your job is actively running—not before or after.
