Instances GCP Compute Engine g4-standard-48 g4-standard-48
g4-standard-48 is a GCP Compute Engine machine type with 48 vCPUs, 180 GiB memory, 1 GPU (AMD EPYC Turin). In Iowa it costs $4.50 per hour on demand ($3,284.93 per month) for Linux; a 1-year committed use discount costs $3.105 per hour, 31% less. Spot capacity is about $1.718 per hour. It is sold in 23 regions.
USD per hour, Linux. Committed use prices are billed monthly. Cheapest on-demand first.
Region On-demand Spot 1-yr committed use 3-yr committed use Iowa us-central1 $4.50 $1.718 $3.105 $1.98 South Carolina us-east1 $4.50 $1.358 $3.105 $1.98 Northern Virginia us-east4 $4.50 $1.813 $3.105 $1.98 Dallas us-south1 $4.50 $1.839 $3.105 $1.98 Oregon us-west1 $4.50 $1.358 $3.105 $1.98 Finland europe-north1 $4.95 $2.097 $3.416 $2.177 Belgium europe-west1 $4.95 $2.222 $3.416 $2.177 Netherlands europe-west4 $4.95 $2.05 $3.416 $2.177 Columbus us-east5 $4.95 $2.522 $3.416 $2.177 Salt Lake City us-west3 $4.95 $2.124 $3.416 $2.177 Las Vegas us-west4 $4.95 $2.135 $3.416 $2.177 Berlin europe-west10 $5.399 $2.234 $3.726 $2.376 Dammam me-central2 $5.399 $2.626 $3.726 $2.376 Singapore asia-southeast1 $5.40 $2.284 $3.726 $2.375 London europe-west2 $5.40 $2.311 $3.726 $2.375 Milan europe-west8 $5.40 $3.04 $3.726 $2.375 Mumbai asia-south1 $5.85 $2.60 $4.038 $2.573 Jakarta asia-southeast2 $5.85 $2.519 $4.175 $2.662 Taiwan asia-east1 $5.85 $2.599 $4.037 $2.573 Tokyo asia-northeast1 $5.85 $2.14 $4.037 $2.573 Delhi asia-south2 $5.85 $2.335 $4.037 $2.573 Melbourne australia-southeast2 $5.85 $2.113 $4.037 $2.573 São Paulo southamerica-east1 $6.434 $3.46 $4.44 $2.83
Compute Value vCPUs 48 Architecture x86_64 Processor AMD EPYC Turin
Memory Value Memory (GiB) 180
Accelerators Value GPUs 1 GPU NVIDIA RTX PRO 6000
Networking Value Network Up to 50 Gbps
Storage Value Storage Network disks only Max persistent disks 32
Instance details Value Instance type g4-standard-48 Family g4 Category Accelerator optimized
Other Value Series G4 Specs source Compute Engine API
Prices updated 2026-11-12 from the Google Cloud Billing Catalog API. Compare all GCP Compute Engine sizes .