Provider workload page

GCP training GPU prices

Training budgets are dominated by 80GB+ accelerators, provider stability, and whether newer H100/H200/B200/MI300X rows justify their premium over A100. The cheapest tracked GCP row in this slice is H100 SXM at $4.84/hr.

80GB+ VRAM on-demand pricing Updated Jun 21, 2026
View qualifying rows All GCP GPUs
Cheapest qualifying row
$4.84/hr
H100 SXM ยท on-demand
Monthly estimate
$3,534/mo
730 hours at current median price
Qualifying GPUs
2
2 current rows updated Jun 21, 2026
Market gap
+$3.84/hr
Floor is Vast.ai A100 PCIE
Live workload pricing

Current GCP rows for training

GCP has 2 qualifying GPU models across 2 current price rows for this workload. The overall tracked market floor is $1.00/hr on Vast.ai A100 PCIE, making GCP $3.84/hr above that floor.

Updated Jun 21, 2026
GPU VRAM Type Median hourly Monthly Offers Updated
H100 SXM
Flagship
 80GB HBM3 on-demand $4.84/hr $3,534/mo 178 Jun 21, 2026
B200
Flagship
 192GB HBM3e on-demand $11.28/hr $8,232/mo 81 Jun 21, 2026
Search guide

GCP training pricing overview

This page narrows GCP's GPU catalog to the rows most likely to matter for training. It is built for buyers who already know the provider they are evaluating and need a workload-specific price shortlist.

Model training, larger fine-tunes, eval farms, and teams comparing H100/A100/H200/B200/MI300X family economics. Use the table below to move from the cheapest visible row into adjacent provider/GPU pages, broader guides, and market-floor comparisons.

Cheapest provider right now

Cheapest GCP row for training

GCP currently starts at $4.84/hr for training with H100 SXM on on-demand pricing. The overall tracked market floor is $1.00/hr on Vast.ai A100 PCIE, making GCP $3.84/hr above that floor.

Methodology and freshness

How this workload slice is computed

We filter the live compare payload to on-demand rows with at least 80GB of VRAM, then sort qualifying GCP rows by current median hourly price.

FAQ

GCP training pricing FAQ

What is the cheapest GCP option for training?

GCP currently starts at $4.84/hr for training with H100 SXM on on-demand pricing.

Is GCP cheapest for training?

The overall tracked market floor is $1.00/hr on Vast.ai A100 PCIE, making GCP $3.84/hr above that floor.

What GPUs count toward this training page?

This page filters to 80GB+ GPUs in the H100 SXM, H100 NVL, H100 PCIE, H200, H200 NVL, and more set and uses on-demand pricing.

How fresh is this GCP training page?

The rows are recalculated from the latest stored provider snapshot. The freshest qualifying row visible here is from Jun 21, 2026.

Popular follow-up searches

Next searches after GCP training

These links move sideways into the full provider catalog, workload guide, market floor, and adjacent provider workload pages.

All GCP GPU prices Open the full provider catalog behind this workload slice. H100 vs A100 for training cost Compare H100-family and A100-family rental costs before choosing a training budget. Market floor: Vast.ai A100 PCIE Compare against the current $1.00/hr training baseline. GCP vLLM hosting prices OpenAI-compatible vLLM serving usually starts at 24GB GPUs and moves into 80GB+ rows for larger models, batching, and long prompts. GCP batch inference prices Batch inference can use cheaper and less stable pricing modes when queues are flexible and retries are acceptable. GCP fine-tuning prices Fine-tuning usually starts in 80GB-class inventory once optimizer state, activations, checkpointing, and sequence length are considered. Best GPU for Qwen Coder 32B hosting See which 48GB-class GPU is the cheapest way to serve Qwen2.5-Coder 32B. Best cloud GPU for fine-tuning Balance 80GB+ memory headroom against hourly spend for adapter and full fine-tunes.