GPU-hour records
Attribute accelerator time to the correct service and customer.
GPU & AI · metering and billing
Turn accelerator consumption into a transparent commercial record.
Measure GPU-hours and supporting services, attribute usage to tenant and project, apply operator-defined plans, and show the customer the same record before invoicing.
- Unit
- GPU-hour
- Resource consumption measured
- Context
- Tenant & project
- Usage has an owner
- Model
- Flexible plans
- On-demand and committed
- Output
- Usage-backed
- Budget, report and invoice
Scope
Metering designed for accelerator products.
GPU usage, reservations and related infrastructure remain connected to the service and plan that explain the charge.
Commitments
Represent reserved and on-demand commercial models.
Supporting usage
Include storage and network consumption around the workload.
Customer transparency
Present project, budget and invoice context in the portal.
Operating workflow
One path from scheduled accelerator to invoice.
Commercial context starts with the service and remains attached through period close.
- 01
Start the record
Attach service, tenant, project and plan at provisioning.
- 02
Measure
Record accelerator and supporting resource consumption.
- 03
Present
Show usage and budget position during the period.
- 04
Reconcile
Apply commitments and issue the usage-backed invoice.
Two connected responsibilities
Operator control and customer action stay explicit.
What commercial teams control
- GPU metering units
- On-demand and commitment plans
- Budget and billing periods
- Invoice and finance exports
What buyers can inspect
- GPU usage by service
- Project and team breakdown
- Commitment and budget position
- Invoices and supporting records
Operational outcome
What this changes in the working model.
Revenue traceability
Every GPU charge points back to delivered capacity.
Commercial flexibility
Shared, reserved and managed products can price differently.
Buyer confidence
Customers see the numbers before invoice close.
Continue in this section
Partner with us
Build the commercial model around every GPU hour.
We will map resource units, commitments, budgets and billing integration into LayerOne.