GCP Compute Instance

n1-t4-2

The n1-t4-2 is a general-purpose GCP Compute instance with 24 vCPUs and 90 GB of memory, 2 T4 GPUs, powered by Intel. Well-suited for web servers, mid-size databases, and backend applications. On-demand pricing starts at $1.8400/hr in us-central1. Spot pricing starts at $0.4203/hr in asia-southeast2.

Price History

Pricing by Region

Region On-Demand ↑Spot Reserved 1Y
us-central1$1.840$1.014—
us-east1$1.840$1.014—
us-west1$1.840$1.014—
europe-west1$1.954$0.931—
europe-west4$1.955$1.166—
asia-east1$2.020$0.714—
us-east4$2.024$0.890—
us-west4$2.024$0.872—
me-west1$2.024$0.968$1.274
northamerica-northeast1$2.025$1.013—
asia-southeast1$2.146$0.533—
asia-south1$2.189$0.758—
us-west2$2.189$0.893—
us-west3$2.189$0.711—
europe-central2$2.199$0.932—
asia-northeast1$2.203$0.946—
asia-northeast3$2.203$1.195—
europe-west2$2.289$0.847—
europe-west3$2.289$1.184—
asia-southeast2$2.339$0.420—
asia-east2$2.475$0.749—
australia-southeast1$2.498$1.228—
southamerica-east1$2.770$1.094—

Spot Interruption Risk

Frequently Asked Questions

What is the n1-t4-2?
The n1-t4-2 is a general-purpose GCP Compute instance with 24 vCPUs and 90 GB of memory, 2 T4 GPUs, powered by Intel. Well-suited for web servers, mid-size databases, and backend applications. On-demand pricing starts at $1.8400/hr in us-central1. Spot pricing starts at $0.4203/hr in asia-southeast2.
What is the cheapest region for n1-t4-2?
us-central1 at $1.840/hr on-demand.
How much does n1-t4-2 cost per month?
Starting at $1343.19/mo on-demand in us-central1, or $306.79/mo with spot pricing.