How much does an idle g4dn.12xlarge cost on AWS?
A g4dn.12xlarge billed on-demand and left running around the clock costs about $2,856/mo at its cheapest listed region (us-east-1). Scheduling it off outside business hours cuts that to roughly $850/mo — a $2,006/mo saving.
Cheapest on-demand
$3.9120/hr
us-east-1
Running 24/7
$2,856/mo
730-hour month
Business-hours savings
$2,006/mo
70% off the 24/7 bill
Price by region
| Region | Hourly (on-demand) | Spot (typical) | ≈ Monthly (24/7) |
|---|---|---|---|
| us-east-1 Cheapest | $3.9120/hr | $1.3692/hr | ≈ $2,856/mo |
| us-east-2 | $3.9120/hr | $1.1345/hr | ≈ $2,856/mo |
| us-west-2 | $3.9120/hr | $1.5648/hr | ≈ $2,856/mo |
| ap-south-1 | $4.3060/hr | $2.0669/hr | ≈ $3,143/mo |
| eu-west-1 | $4.3620/hr | $1.8320/hr | ≈ $3,184/mo |
| eu-central-1 | $4.8900/hr | $1.6626/hr | ≈ $3,570/mo |
| ap-southeast-2 | $5.0870/hr | $1.9331/hr | ≈ $3,714/mo |
| ap-northeast-1 | $5.2810/hr | $2.5349/hr | ≈ $3,855/mo |
| ap-southeast-1 | $5.4740/hr | $2.1349/hr | ≈ $3,996/mo |
| sa-east-1 | $6.6490/hr | $2.0612/hr | ≈ $4,854/mo |
What a schedule saves
| Schedule | Monthly cost | Savings vs 24/7 |
|---|---|---|
| Running 24/7 | $2,856/mo | baseline |
| Weekdays only (off on weekends) | $2,040/mo | -$816/mo (29%) |
| Business hours only (10×5) | $850/mo | -$2,006/mo (70%) |
Assumes a 730-hour month at on-demand list price. Actual savings depend on when the instance is actually used.
About the g4dn family
G4dn pairs NVIDIA T4 GPUs with Intel Cascade Lake vCPUs, the workhorse choice for ML inference, small-scale training, video transcoding, and cloud gaming or virtual workstations. It's also the default GPU dev box: a data scientist spins one up to prototype a model, then it sits idle every night and weekend while still billing for the GPU — the single most expensive kind of forgotten instance we see.
Regional pricing varies: g4dn.12xlarge runs $3.9120/hr in us-east-1, the cheapest region priced here, versus $6.6490/hr in sa-east-1, the priciest — a 41% spread across regions for the identical instance type.
The next size up in the same family, g4dn.16xlarge (64 vCPU / 256 GB RAM), costs about $3,177/mo running 24/7 at its cheapest region — roughly 11% more than g4dn.12xlarge's ~$2,856/mo — for 1.3x the vCPU. Note the trade: it carries 1 GPU versus 4 on the g4dn.12xlarge — a trade that only makes sense if your work isn't GPU-bound.
Common questions
How much does an idle g4dn.12xlarge cost per month?
At the cheapest listed region (us-east-1), a g4dn.12xlarge costs $3.9120/hr, or about $2,856/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 $2,006/mo (70%) versus running 24/7. A lighter schedule — on around the clock on weekdays, off on weekends — still saves about $816/mo (29%).
Do I still pay for the GPU when a g4dn instance is stopped?
No — a stopped instance accrues no compute charges at all, GPU included; you keep paying only for its EBS volumes. That is exactly why the lease model pays off fastest on GPU boxes: the hardware premium disappears the moment the machine stops, and the environment survives on its EBS volumes — just don't stage anything you need on the local NVMe scratch drives, which do not survive a stop.
Other g4dn sizes
Stop paying for idle AWS 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 Linux 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 2026-07-20. 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.