Hyperscaler · 8 GPU models tracked
Google Cloud
$0.16–$16.11per GPU-hour, the cheapest row to the dearest
GPU rates in US regions, the cheapest region named in each row: on-demand, spot, and 1- and 3-year commitments. The rate is the GPU alone - the VM's CPU and memory are billed on top, so a running instance costs more.
As of , Google Cloud lists 8 GPU models we track, from $0.16 per GPU-hour (NVIDIA T4, 3-year, GPU only) to $16.11 (NVIDIA B200, on-demand, GPU only).
- Cheapest row$0.16NVIDIA T4 · 3-year, GPU only
- B200$16.11on-demand · us-west1, GPU only · 132% above the median $6.94 (40) · spot from $1.45
- H200$9.31on-demand · us-west1, GPU only · 104% above the median $4.56 (56) · spot from $5.27
- H100 SXM$9.80on-demand · us-west8, GPU only · 198% above the median $3.29 (41) · spot from $1.06
| Website | cloud.google.com |
|---|---|
| Pricing page | cloud.google.com/products/compute/pricing/accelerator-optimized |
| Regions and tiers | us-central1, us-east4, us-east5, us-west1, us-west2, us-west4, us-west8 — as labelled in the rates below |
| Models tracked | 8 |
Price by GPU
As of , per GPU-hour: Google Cloud’s on-demand rate for each card it lists, a cheaper market it also sells, and where the on-demand rate stands among the providers we track.
| GPU | Google Cloud on-demand, per GPU-hour | Against the market (median rate card) |
|---|---|---|
| NVIDIA B200 | $16.11 us-west1, GPU only captured spot $1.45 Deploy ↗ | 132% above the median $6.94 · 40 providers 39 of 40 list lower · lowest Hostinger $4.50 from · compare |
| NVIDIA H100 SXM | $9.80 us-west8, GPU only captured spot $1.06 Deploy ↗ | 198% above the median $3.29 · 41 providers 38 of 41 list lower · lowest Cudo Compute $1.79 GPU only · compare |
| NVIDIA H200 | $9.31 us-west1, GPU only captured spot $5.27 Deploy ↗ | 104% above the median $4.56 · 56 providers 52 of 56 list lower · lowest Dataoorts $2.29 · compare |
| NVIDIA A100 80GB | $3.93 us-central1, GPU only captured spot $0.40 Deploy ↗ | 97% above the median $1.99 · 63 providers 54 of 63 list lower · lowest Yotta Labs $0.92 · compare |
| NVIDIA A100 40GB | $2.93 us-central1, GPU only captured spot $0.29 Deploy ↗ | 47% above the median $1.99 · 22 providers 17 of 22 list lower · lowest Trooper.AI $0.69 from · compare |
| NVIDIA V100 | $2.48 us-west4, GPU only captured Deploy ↗ | 199% above the median $0.83 · 13 providers 7 of 13 list lower · lowest Verda $0.24 · compare |
| NVIDIA L4 | $0.56 us-east4, GPU only captured Deploy ↗ | 30% below the median $0.80 · 23 providers 4 of 23 list lower · lowest Hikube $0.28 GPU only · compare |
| NVIDIA T4 | $0.35 us-east5, GPU only captured Deploy ↗ | 36% below the median $0.55 · 19 providers 2 of 19 list lower · lowest immers.cloud $0.24 · compare |
The same, in words
- Google Cloud lists the NVIDIA B200 from $16.11 per GPU-hour on-demand (us-west1, GPU only, same rate in all 3 regions read), $1.45 spot (us-east5, GPU only), $7.09 reserved (3yr, same rate in all 7 regions read). The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 132% above the median rate card ($6.94 across 40 providers); 39 of the 40 providers we track list a lower on-demand rate, the lowest Hostinger at $4.50 (Hostinger’s rates, Google Cloud vs Hostinger).
- Google Cloud lists the NVIDIA H100 SXM from $9.80 per GPU-hour on-demand (us-west8, GPU only, cheapest of 10 regions read; dearest $11.78), $1.06 spot (us-west2, GPU only), $4.30 reserved (3yr, cheapest of 10 regions read; dearest $5.17). The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 198% above the median rate card ($3.29 across 41 providers); 38 of the 41 providers we track list a lower on-demand rate, the lowest Cudo Compute at $1.79 (Cudo Compute’s rates, Google Cloud vs Cudo Compute).
- Google Cloud lists the NVIDIA H200 from $9.31 per GPU-hour on-demand (us-west1, GPU only, same rate in all 4 regions read), $5.27 spot (us-east4, GPU only), $4.08 reserved (3yr, cheapest of 5 regions read; dearest $4.82). The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 104% above the median rate card ($4.56 across 56 providers); 52 of the 56 providers we track list a lower on-demand rate, the lowest Dataoorts at $2.29 (Dataoorts’s rates, Google Cloud vs Dataoorts).
- Google Cloud lists the NVIDIA A100 80GB from $3.93 per GPU-hour on-demand (us-central1, GPU only, cheapest of 10 regions read; dearest $4.72), $0.40 spot (us-west4, GPU only), $3.10 reserved (3yr, cheapest of 9 regions read; dearest $3.72). The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 97% above the median rate card ($1.99 across 63 providers); 54 of the 63 providers we track list a lower on-demand rate, the lowest Yotta Labs at $0.92 (Yotta Labs’s rates, Google Cloud vs Yotta Labs).
- Google Cloud lists the NVIDIA A100 40GB from $2.93 per GPU-hour on-demand (us-central1, GPU only, cheapest of 10 regions read; dearest $3.46), $0.29 spot (us-west8, GPU only), $1.03 reserved (3yr, cheapest of 10 regions read; dearest $1.21). The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 47% above the median rate card ($1.99 across 22 providers); 17 of the 22 providers we track list a lower on-demand rate, the lowest Trooper.AI at $0.69 (Trooper.AI’s rates, Google Cloud vs Trooper.AI).
- Google Cloud lists the NVIDIA V100 from $2.48 per GPU-hour on-demand (us-west4, GPU only, cheapest of 8 regions read; dearest $2.98), $1.12 reserved (3yr, cheapest of 8 regions read; dearest $1.34). The on-demand rate is unchanged since first read, Sep 23, 2026 (7-day change: none). Against the market: 199% above the median rate card ($0.83 across 13 providers); 7 of the 13 providers we track list a lower on-demand rate, the lowest Verda at $0.24 (Verda’s rates, Google Cloud vs Verda).
- Google Cloud lists the NVIDIA L4 from $0.56 per GPU-hour on-demand (us-east4, GPU only, cheapest of 10 regions read; dearest $0.67), $0.25 reserved (3yr, cheapest of 10 regions read; dearest $0.30). The on-demand rate is unchanged since first read, Sep 23, 2026 (7-day change: none). Against the market: 30% below the median rate card ($0.80 across 23 providers); 4 of the 23 providers we track list a lower on-demand rate, the lowest Hikube at $0.28 (Hikube’s rates, Google Cloud vs Hikube).
- Google Cloud lists the NVIDIA T4 from $0.35 per GPU-hour on-demand (us-east5, GPU only, cheapest of 10 regions read; dearest $0.41), $0.16 reserved (3yr, cheapest of 10 regions read; dearest $0.19). The on-demand rate is unchanged since first read, Sep 23, 2026 (7-day change: none). Against the market: 36% below the median rate card ($0.55 across 19 providers); 2 of the 19 providers we track list a lower on-demand rate, the lowest immers.cloud at $0.24 (immers.cloud’s rates, Google Cloud vs immers.cloud).
All listed prices
One table per price type, never mixed. Every rate normalized to one GPU for one hour; a price opens the page it was read from, and 7d marks a rate that moved in the past seven days. The day each figure was first listed is in this page’s data, below.
On-demand
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| NVIDIA T4GPU onlyAPIus-east5 | $0.35 | captured | Deploy ↗ |
| NVIDIA L4GPU onlyAPIus-east4 | $0.56 | captured | Deploy ↗ |
| NVIDIA V100GPU onlyAPIus-west4 | $2.48 | captured | Deploy ↗ |
| NVIDIA A100 40GBGPU onlyAPIus-central1 | $2.93 | captured | Deploy ↗ |
| NVIDIA A100 80GBGPU onlyAPIus-central1 | $3.93 | captured | Deploy ↗ |
| NVIDIA H200GPU onlyAPIus-west1 | $9.31 | captured | Deploy ↗ |
| NVIDIA H100 SXMGPU onlyAPIus-west8 | $9.80 | captured | Deploy ↗ |
| NVIDIA B200GPU onlyAPIus-west1 | $16.11 | captured | Deploy ↗ |
Spot / interruptible
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| NVIDIA A100 40GBGPU onlyAPIus-west8 · 7d −5.0% | $0.29 | captured | Deploy ↗ |
| NVIDIA A100 40GBGPU onlyAPIus-west2 | $0.31 | stoppedlast confirmed | |
| NVIDIA A100 80GBGPU onlyAPIus-west4 | $0.40 | captured | Deploy ↗ |
| NVIDIA H100 SXMGPU onlyAPIus-west2 | $1.06 | captured | Deploy ↗ |
| NVIDIA B200GPU onlyAPIus-east5 | $1.45 | captured | Deploy ↗ |
| NVIDIA H200GPU onlyAPIus-east4 · 7d −5.0% | $5.27 | captured | Deploy ↗ |
Reserved (committed use)
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| NVIDIA T4GPU onlyAPI3-year · us-west8 · min. commitment 3yr | $0.16 | captured | Reserve ↗ |
| NVIDIA T4GPU onlyAPI1-year · us-east5 · min. commitment 1yr · −$0.13 (−37%) vs its on-demand $0.35 | $0.22 | captured | Reserve ↗ |
| NVIDIA L4GPU onlyAPI3-year · us-east4 · min. commitment 3yr · −$0.31 (−55%) vs its on-demand $0.56 | $0.25 | captured | Reserve ↗ |
| NVIDIA L4GPU onlyAPI1-year · us-east4 · min. commitment 1yr · −$0.21 (−37%) vs its on-demand $0.56 | $0.35 | captured | Reserve ↗ |
| NVIDIA A100 40GBGPU onlyAPI3-year · us-west8 · min. commitment 3yr | $1.03 | captured | Reserve ↗ |
| NVIDIA V100GPU onlyAPI3-year · us-central1 · min. commitment 3yr | $1.12 | captured | Reserve ↗ |
| NVIDIA V100GPU onlyAPI1-year · us-west8 · min. commitment 1yr | $1.56 | captured | Reserve ↗ |
| NVIDIA A100 40GBGPU onlyAPI1-year · us-central1 · min. commitment 1yr · −$1.09 (−37%) vs its on-demand $2.93 | $1.85 | captured | Reserve ↗ |
| NVIDIA A100 80GBGPU onlyAPI3-year · us-central1 · min. commitment 3yr · −$0.83 (−21%) vs its on-demand $3.93 | $3.10 | captured | Reserve ↗ |
| NVIDIA A100 80GBGPU onlyAPI1-year · us-central1 · min. commitment 1yr · −$0.45 (−11%) vs its on-demand $3.93 | $3.48 | captured | Reserve ↗ |
| NVIDIA H200GPU onlyAPI3-year · us-west1 · min. commitment 3yr · −$5.23 (−56%) vs its on-demand $9.31 | $4.08 | captured | Reserve ↗ |
| NVIDIA H100 SXMGPU onlyAPI3-year · us-east4 · min. commitment 3yr | $4.30 | captured | Reserve ↗ |
| NVIDIA H200GPU onlyAPI1-year · us-east4 · min. commitment 1yr | $6.43 | captured | Reserve ↗ |
| NVIDIA H100 SXMGPU onlyAPI1-year · us-central1 · min. commitment 1yr | $6.80 | captured | Reserve ↗ |
| NVIDIA B200GPU onlyAPI3-year · us-west2 · min. commitment 3yr | $7.09 | captured | Reserve ↗ |
| NVIDIA B200GPU onlyAPI1-year · us-west1 · min. commitment 1yr · −$4.99 (−31%) vs its on-demand $16.11 | $11.12 | captured | Reserve ↗ |
Ask an AI agent about Google Cloud’s rates: Claude, Cursor, VS Code and ChatGPT can read them through the GPU Economy MCP server - free, no key, every answer citing the page and read time.
What else Google Cloud bills
The GPU-hour is not the whole bill. What Google Cloud's own pages say about the rest, read by hand and checked on ; each line links the page it was read from. "Not published" means we looked and the pages do not say. Every provider side by side.
| What | What Google Cloud states | Source |
|---|---|---|
| In the rate | Listed A3/A2 price includes GPUs, vCPU, memory and bundled Local SSD (a3-highgpu-8g: 208 vCPU, 1,871 GB, 6,000 GiB SSD); boot disk extra | source checked |
| Data out | Premium Tier (default), us-central1 to North America: first 1 GiB free, then $0.12/GiB up to 1 TiB | source checked |
| Storage | Balanced Persistent Disk $0.000136986/GiB-hour in us-central1, about $0.10/GiB-month; Hyperdisk Balanced about $0.08 plus IOPS/throughput our reading | source checked |
| While stopped | Stopped instance: no vCPU/memory charge; disks and external IPs billed regardless of state | source checked |
| Billing step | Per second after a 1-minute minimum (vCPUs, GPUs, memory) | source checked |
Every rate on this page as data: /providers/gcp/prices.json — the same fields as prices.json, narrowed to this provider. CC BY 4.0, no key, no rate limit. On your own page: embed this table (/embed/providers/gcp, an iframe, no script).
Compare Google Cloud with
Every provider that lists at least two of the same GPUs on-demand, most shared GPUs first; each opens the side-by-side page.
- Amazon Web Services 8 shared GPUs
- EmpirioLabs 7 shared GPUs
- Modal 7 shared GPUs
- RunPod 7 shared GPUs
- Vast.ai 7 shared GPUs
- Cerebrium 6 shared GPUs
- E2E Networks 6 shared GPUs
- Microsoft Azure 6 shared GPUs
- Oracle Cloud (OCI) 6 shared GPUs
- Sesterce 6 shared GPUs
- Verda 6 shared GPUs
- GPU.ai 5 shared GPUs
- Hugging Face 5 shared GPUs
- Jarvislabs 5 shared GPUs
- Lambda 5 shared GPUs
- Lightning AI 5 shared GPUs
- Northflank 5 shared GPUs
- Barrack AI 4 shared GPUs
- Baseten 4 shared GPUs
- Civo 4 shared GPUs
- CoreWeave 4 shared GPUs
- HexGrid 4 shared GPUs
- Hyperstack 4 shared GPUs
- Koyeb 4 shared GPUs
- Lium 4 shared GPUs
- Paperspace 4 shared GPUs
- Seeweb 4 shared GPUs
- Beam 3 shared GPUs
- Bitdeer AI 3 shared GPUs
- Cloud.ru Evolution 3 shared GPUs
- Crusoe 3 shared GPUs
- Deep Infra 3 shared GPUs
- Denvr Dataworks 3 shared GPUs
- Elice Cloud 3 shared GPUs
- FarmGPU 3 shared GPUs
- Gcore 3 shared GPUs
- Geodd 3 shared GPUs
- Hikube 3 shared GPUs
- HPC-AI.COM 3 shared GPUs
- Hyperbolic 3 shared GPUs
- immers.cloud 3 shared GPUs
- Lyceum 3 shared GPUs
- Nebius 3 shared GPUs
- Neysa 3 shared GPUs
- OpenRelay 3 shared GPUs
- OVHcloud 3 shared GPUs
- SynpixCloud 3 shared GPUs
- Tata Communications 3 shared GPUs
- Together AI 3 shared GPUs
- Yotta Labs 3 shared GPUs
- airgpu 2 shared GPUs
- Atlas Cloud 2 shared GPUs
- Bamboozle Web Services 2 shared GPUs
- Canopy Wave 2 shared GPUs
- Cato Digital 2 shared GPUs
- centron 2 shared GPUs
- CloudGPU 2 shared GPUs
- CloudRift 2 shared GPUs
- Cudo Compute 2 shared GPUs
- Dataoorts 2 shared GPUs
- Daytona 2 shared GPUs
- fal 2 shared GPUs
- Fireworks AI 2 shared GPUs
- FlexAI 2 shared GPUs
- FluidCore 2 shared GPUs
- FPT AI Factory 2 shared GPUs
- GMI Cloud 2 shared GPUs
- Hostinger 2 shared GPUs
- Inferless 2 shared GPUs
- Infomaniak 2 shared GPUs
- Limestone Networks 2 shared GPUs
- Massed Compute 2 shared GPUs
- packet.ai 2 shared GPUs
- Replicate 2 shared GPUs
- RunComfy 2 shared GPUs
- Runware 2 shared GPUs
- Selectel 2 shared GPUs
- Servercore 2 shared GPUs
- STACKIT 2 shared GPUs
- Taiga Cloud (Northern Data Group) 2 shared GPUs
- Thunder Compute 2 shared GPUs
- Trooper.AI 2 shared GPUs
- VESSL AI 2 shared GPUs
- Yandex Cloud 2 shared GPUs
Data updated: