A2 family · Accelerator Optimizedx86_64 · Intel/AMD

a2-highgpu-1g

The a2-highgpu-1g machine type has 12 vCPUs and 85 GB of memory. Pricing for this instance starts at $3.67 per hour and $2681.57 monthly in the us-central1 region.

Updated June 26, 2026
vCPUs
12
Memory
85 GB
Network
Up to 24 Gbps
Storage
Local SSD

Save up to $1341/mo on every a2-highgpu-1g

Connect Google Cloud — we'll find the optimal pricing model for your workload.

All pricing options

us-central1 · USD · monthly = 730 hrs
Prices exclude local taxes
TermHourlyMonthlySavings
On-Demand
$3.6734$2681.57 Baseline
1-Year Committed Use Discount (CUD)
$2.3142$1689.37 -37%
3-Year Committed Use Discount (CUD)
$1.2857$938.57 -65%
Preemptible / Spot
$1.8363$1340.54 -50%
You could save $1341.04/mo on this instance with the right pricing model.
Find my savings

Monthly cost estimator

Estimate your spend based on actual usage.
730 h
1h365h730h (24/7)
1
Estimated monthly
$2681.57
730 hrs × $3.6734/hr × 1

Compute

x86_64 · Intel/AMD
vCPUs i12
Memory i85 GB
Physical processorIntel Cascade Lake
Nested VirtualizationNot supported
Sole TenantNot supported
GPU1 GPU
Within A2 family
a2-highgpu-1g
12 vCPU
a2-ultragpu-1g
12 vCPU
a2-highgpu-2g
24 vCPU
a2-ultragpu-2g
24 vCPU
a2-highgpu-4g
48 vCPU
a2-ultragpu-4g
48 vCPU
a2-highgpu-8g
96 vCPU
a2-ultragpu-8g
96 vCPU
a2-megagpu-16g
96 vCPU

Lock in this rate across your fleet — typically save 30–40%

Connect your Google Cloud account to apply this pricing logic to every running instance automatically.

Networking

Max egress bandwidth24 Gbps
Tier 1 bandwidthN/A
Enhanced networking iYes (gvnic/virtio)
IPv6 supportYes (Dual-stack VPC)

Storage

Max persistent disks128
Max disk size257 TB
Local SSDsSupported (Scratch Disk)

Compare with another instance

vs.
Speca2-highgpu-1ga2-ultragpu-1gΔ
vCPUs12120%
Memory85 GB170 GB+100%
Hourly Price$3.6734$5.0688+38%
Monthly Price$2681.57$3700.22+38%
Cost optimization

GCP recommender alerts can cut this bill ~30-40%.

Connect your GCP account. We'll identify every underutilized a2.* instance and show exactly where downsizing, cleaning up orphaned disks, or shifting to CUDs saves money — with zero code changes.