Usage and billing
Read the Billing tab on the project Summary page - what is metered, the units used, and how to download a usage report.
Each project's Summary page has a Billing tab that shows metered usage for the project. The console shows usage quantities only — prices and invoices come from your provider or administrator.
How metering works
CPU and memory are metered by allocation, not by actual consumption: what counts is what your resources are configured with, sampled at a regular interval (every few minutes) while the resource exists. Disk is the one exception: storage volumes are metered by their configured size, but an instance's root disk is metered by its actual usage. A stopped instance with limits.cpu: 4 accrues the same core-minutes as a busy one.
What each resource kind reports:
| Resource | Metered as |
|---|---|
| Instances | CPU from limits.cpu, memory from limits.memory, disk from the root disk |
| Storage volumes | Configured volume size |
| Akasha PSQL databases | Requested CPU, memory, and configured storage size, each multiplied by the number of database instances |
| Managed clusters | A flat per-minute service fee per provisioned cluster (the cluster's node VMs are billed as instances) |
Network traffic and GPUs are not metered.
Units
| Unit | Meaning |
|---|---|
| Core-minutes | Allocated CPU cores × minutes |
| RAM GB-minutes | Allocated memory in GB × minutes (1 GB here = 1024³ bytes) |
| Disk / Storage GB-minutes | Disk in GB × minutes (configured size for volumes, actual root-disk usage for instances) |
| DB core / RAM GB / Storage GB-minutes | The same units for Akasha PSQL databases |
| Managed cluster minutes | Minutes each managed Kubernetes cluster existed |
Read the Billing tab
- Go to your project's Summary page and open the Billing tab.
- Pick a period: This month, Last month, Year to date, or a specific month from the month picker.
- Review the totals, grouped as:
- Machine compute — core-minutes, RAM GB-minutes, and disk/storage GB-minutes for instances and volumes.
- Database compute (Akasha PSQL) — the database equivalents.
- Platform service fee — managed cluster minutes, a per-cluster fee separate from the underlying machines.
The tab refreshes automatically while open. Viewing a project's usage requires access to that project; the aggregate view across all projects is administrator-only.
Per-resource details
Below the totals, usage is broken down per resource, grouped by kind:
- Instances — one row per instance.
- Storage volumes — one row per volume, with a Purpose column that identifies Kubernetes control-plane and worker-pool disks (with the cluster name) versus custom volumes.
- Akasha PSQL databases — one row per database.
- Managed clusters — one row per cluster.
- Virtual machines (KubeVirt) — VMs running inside your Kubernetes clusters, marked "Visibility only · not added to totals": their CPU and memory already run on cluster node VMs that are billed as instances, so these rows are informational and are not counted in the totals.
Each row shows the resource's usage and its active minutes in the period. Resources that no longer exist keep their accrued usage and are marked with a "deleted" tag.
Download a usage report
Click Download PDF on the Billing tab to save a usage report for the selected period.
Prices and invoices
Akasha meters quantities; it does not generate invoices or show prices to project users. Rates and billing arrangements are set by your provider — contact your administrator for pricing questions.
If the Billing tab shows no data or an error, usage metering may not be enabled in your deployment — ask your administrator.
Usage via the API
The same data is available at GET /1.0/billing/usage?project=<project>&start=<RFC3339>&end=<RFC3339> (omit start and end for the current month). See API endpoints.