MACHINE AGE AI — GPU QUOTE CHECKLIST
Version: 30 September 2026
https://machineage.ai/guides/compare-gpu-rental-quotes

WORKLOAD
Task/model:
Required GPU model or acceptable alternatives:
GPU count and minimum memory per GPU:
CPU/RAM/local disk requirements:
Region and any data-location requirement:
Start date/time/time zone:
Expected runtime and maximum billable hours:
Interconnect/multi-node requirement:
Software/container and access method:
Checkpoint/restart tolerance:

COST AND TERMS
Is the quoted rate per GPU-hour or per complete instance-hour?
GPU count included in the quoted instance:
Compute rate and currency:
Storage while running and stopped:
Expected ingress/egress and bandwidth rate:
Setup, minimum-duration, support and other fees:
Tax treatment and total before/after tax:
Quote expiry and cancellation terms:
What can be interrupted, and with what notice?

CONFIRMATION AND DELIVERY
Who contracts, invoices and supports the workload?
What hardware, region and start time are confirmed?
How will access be delivered and tested?
What service limits and remedy are agreed?
Owner authorized to approve the purchase:
Maximum authorized total:

Estimated total = hourly rate × billable hours × applicable GPU count
                  + storage + transfer + other charges + applicable tax.
Do not multiply a complete-instance rate by GPU count again.
Compare cost per completed job where measured throughput is available.

Send a non-binding requirement: https://machineage.ai/market/compute
A listing or estimate is not a reservation, supplier confirmation or delivery.