Prices updated 2026-06-23 — sourced from live provider data

Google Cloud H200 Pricing Options

All available H200 instance types on Google Cloud, sorted by lowest hourly rate.

Region Instance Type Price/hr GPUs vCPUs RAM
us-central1 a3-megagpu-8g (8x H200) On-Demand $40.3200/hr Cheapest 8× GPU 208 vCPU 1872 GB

About Google Cloud H200 Cloud Pricing

Google Cloud Platform offers GPU compute through its A2, A3, and G2 machine series, with deep integration into Vertex AI, TPU pods, and Google's proprietary ML infrastructure.

Why Choose Google Cloud for H200

  • TPU v5p available alongside GPU options for specific workloads
  • Vertex AI for managed ML pipelines
  • Sustained use discounts (automatic, no commitment needed)
  • Preemptible VMs for up to 60–91% savings on batch jobs

Best suited for: Teams using TensorFlow/JAX who want access to both GPU and TPU hardware, plus managed ML services via Vertex AI.

H200 Technical Overview

ArchitectureHopper (Enhanced)
Memory141 GB HBM3e
Best ForLarge-context inference, memory-bound training, and workloads that exceed H100's 80 GB VRAM limit

Google Cloud (GCP) offers H100 via their A3 Ultra machines in us-central1, us-east4, and europe-west4 -- exclusively as on-demand instances with sustained-use discounts applied automatically (no commitment required). GCP does not offer H100 spot instances, and reserved pricing requires 1 or 3-year committed use contracts. A key differentiator: GCP's H100 VMs come with 400 Gbps InfiniBand (via TrDeal links) enabling multi-node training clusters with near-NVSwitch-equivalent bandwidth, and you can attach Google Cloud's TPU v5p pods alongside H100 VMs for heterogeneous training pipelines. GCP's H100 pricing typically runs 15-25% above Lambda/CoreWeave equivalent configs.

Pricing Analysis

Google Cloud's cheapest H200 option starts at $40.3200/hr, which is 877% above the market average of $4.1287/hr. Google Cloud currently offers on-demand pricing only for H200. This means no commitment required, but you won't access the deeper discounts available from reserved or spot pricing.

Note: GCP's sustained use discounts provide automatic savings without commitments, but their GPU availability can be more constrained than AWS in certain regions.

Market context: H200 availability is growing rapidly in 2026 as providers upgrade their H100 clusters. Pricing typically runs 20–40% above H100 equivalent configurations.

Top Alternatives to Google Cloud for H200

Side-by-side cost comparison vs the cheapest Google Cloud H200 rate ($40.3200/hr).

Provider Price/hr Type vs Google Cloud
RunPod $0.5000/hr On-Demand -99% cheaper vs Google Cloud
Lambda $2.9900/hr On-Demand -93% cheaper vs Google Cloud
Crusoe Energy $3.1500/hr On-Demand -92% cheaper vs Google Cloud
Crusoe Cloud $3.6000/hr On-Demand -91% cheaper vs Google Cloud
Nebius $3.6000/hr On-Demand -91% cheaper vs Google Cloud

View all H200 providers →

Calculate Your Exact H200 Cost

Enter your hours, job type, and GPU count. Our calculator compares Google Cloud against all 32 providers instantly.

● $19 — no subscription
Track Google Cloud H200 prices — daily alerts
Get emailed when Google Cloud drops below your budget. $19 for 30 days. No account needed.
Set Price Alert → Calculator →

Frequently Asked Questions

How much does Google Cloud H200 cost per hour in 2026? +
Google Cloud H200 starts at $40.3200/hr (on-demand). The market average for H200 across all providers is $4.1287/hr. Use the GPU cost calculator to estimate total cost based on your hours and workload type.
What are the cheapest alternatives to Google Cloud for H200? +
There are 9 providers offering H200 cheaper than Google Cloud: RunPod at $0.5000/hr, Lambda at $2.9900/hr, Crusoe Energy at $3.1500/hr, and 6 more. 3 other providers offer H200 at higher rates. See the full comparison table for all H200 providers.
Does Google Cloud offer reserved or spot H200 pricing? +
Google Cloud offers 1 pricing type for H200: on-demand. With on-demand only, you pay per hour with no commitment. While you won't get reserved or spot discounts, this provides maximum flexibility for unpredictable workloads.
What workloads is Google Cloud H200 best suited for? +
The H200 (Hopper (Enhanced) architecture, 141 GB HBM3e) is best suited for Large-context inference, memory-bound training, and workloads that exceed H100's 80 GB VRAM limit. Google Cloud is a hyperscaler provider — Teams using TensorFlow/JAX who want access to both GPU and TPU hardware, plus managed ML services via Vertex AI.
How much does it cost to run H200 on Google Cloud for a month? +
At Google Cloud's cheapest H200 rate of $40.3200/hr: running 24/7 costs ~$967.68/day or ~$29433.60/month per GPU. For training jobs, costs scale linearly — a 100-hour training run would cost ~$4032.00. Multi-GPU configurations multiply accordingly. Use the cost calculator for precise estimates with your specific parameters.
Compare across all providers

Not sure which GPU configuration fits your budget?

Get a personalized recommendation based on your monthly budget, workload, and real pricing data from all 58 providers.

Budget Optimizer — $19 →
$19 One-Time

Is a reserved contract worth it here?

Get the break-even math for Google Cloud H200 — calculated from real 90-day pricing data, not an LLM.

Get Analysis — $19