How much does an idle a3-ultragpu-8g cost on Google Cloud?
An a3-ultragpu-8g billed on-demand and left running around the clock costs about $60,949/mo at its cheapest listed region (us-west1). Scheduling it off outside business hours cuts that to roughly $18,140/mo — a $42,809/mo saving.
Cheapest on-demand
$83.4918/hr
us-west1
Running 24/7
$60,949/mo
730-hour month
Business-hours savings
$42,809/mo
up to 70% off the 24/7 bill (est.)
Price by region
| Region | Hourly (on-demand) | Spot (typical) | ≈ Monthly (24/7) |
|---|---|---|---|
| us-west1 Cheapest | $83.4918/hr | $48.2176/hr | ≈ $60,949/mo |
| europe-west1 | $91.9472/hr | $55.1555/hr | ≈ $67,121/mo |
What a schedule saves
| Schedule | Monthly cost | Savings vs 24/7 |
|---|---|---|
| Running 24/7 | $60,949/mo | baseline |
| Weekdays only (off on weekends) | $43,535/mo | -$17,414/mo (29%) |
| Business hours only (10×5) | $18,140/mo | -$42,809/mo (70%) |
Assumes a 730-hour month at on-demand list price. Actual savings depend on when the instance is actually used.
About the a3-ultra family
A3 Ultra is GCP's H200 platform: eight GPUs with 141 GB of memory each on a host built for multi-node training, where memory per GPU and the interconnect set the pace. It is sold in a short list of regions and often through reservations, so teams that win capacity hold on to it — and a held machine bills the same whether a job is running or the team is still deciding what to run next.
Regional pricing varies: a3-ultragpu-8g runs $83.4918/hr in us-west1, the cheapest region priced here, versus $91.9472/hr in europe-west1, the priciest — a 9% spread across regions for the identical instance type.
Common questions
How much does an idle a3-ultragpu-8g cost per month?
At the cheapest listed region (us-west1), an a3-ultragpu-8g costs $83.4918/hr, or about $60,949/mo if it runs 24/7. That's the on-demand list price — an idle instance costs exactly the same as a busy one, since cloud billing doesn't know the difference.
How much can I save by scheduling it off outside work hours?
Restricting it to business hours only (10×5) saves about $42,809/mo (70%) versus running 24/7. A lighter schedule — on around the clock on weekdays, off on weekends — still saves about $17,414/mo (29%).
Why does the H200 show up in so few GCP regions?
Because A3 Ultra is capacity-constrained hardware: Google sells it in a handful of zones, and in some of them only through reservations or with an account team involved. The table shows on-demand list rates only where the machine type is actually purchasable that way. For teams holding a reservation, an idle hour is already paid for — the schedule argument applies to on-demand capacity.
Stop paying for idle Google Cloud servers
Idlefy runs a free audit of your fleet, then flips dev servers to stopped-by-default: lease one when you need it, and
when the lease ends it shuts itself off — stop, not terminate, persistent disks intact. Timers are set by your team, never
guessed, and Idlefy only manages instances tagged idlefy=enabled — on AWS that boundary is enforced by the IAM policy itself.
Prices are on-demand list price for standard instances and exclude storage, network, data transfer, and any negotiated or volume discounts. Actual bills vary by usage and account-level pricing. Prices as of 2026-08-31. Spot figures are indicative — AWS values are derived from AWS's published typical savings vs on-demand for the region; GCP values are the listed Spot price. Spot capacity can be reclaimed by the provider. Stopping preserves EBS and persistent-disk volumes; data on local instance-store or local-SSD drives does not survive a stop.