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

RunPod H100 Pricing Options

All available H100 instance types on RunPod, sorted by lowest hourly rate.

Region Instance Type Price/hr GPUs vCPUs RAM
US/EU H100 SXM (Community) Spot $1.8900/hr Cheapest 1× GPU 16 vCPU 251 GB
community NVIDIA H100 PCIe-spot Spot $1.9900/hr 1× GPU
us-east-1 NVIDIA H100 PCIe On-Demand $1.9900/hr 1× GPU
US/EU H100 SXM 80GB On-Demand $2.4900/hr 1× GPU 16 vCPU 251 GB
us-east-1 NVIDIA H100 NVL On-Demand $2.5900/hr 1× GPU
community NVIDIA H100 NVL-spot Spot $2.5900/hr 1× GPU
community NVIDIA H100 80GB HBM3-spot Spot $2.6900/hr 1× GPU
us-east-1 NVIDIA H100 80GB HBM3 On-Demand $2.6900/hr 1× GPU

About RunPod H100 Cloud Pricing

RunPod operates a GPU cloud marketplace that aggregates capacity from multiple data centers, offering both community cloud (lower-cost shared infrastructure) and secure cloud (dedicated servers) tiers.

Why Choose RunPod for H100

  • Among the lowest on-demand GPU prices in the market
  • Serverless GPU endpoints for inference workloads
  • Wide GPU model selection from consumer (RTX 3090) to enterprise (H100)
  • Template marketplace for one-click ML environment deployment

Best suited for: Individual developers, hobbyists, and cost-sensitive teams who prioritize price over enterprise features. RunPod's community cloud tier is often the cheapest option for experimentation.

H100 Technical Overview

ArchitectureHopper
Memory80 GB HBM3
Best ForLarge language model training, foundation model pre-training, and high-throughput inference

RunPod offers H100 in both their Secure Cloud (dedicated bare-metal) and Community Cloud (shared) tiers -- the only provider in the top 5 with genuine spot-style H100 pricing via their interruptible Secure Cloud option. H100 availability fluctuates based on demand and data center capacity, which RunPod's API exposes in real-time (check /v2/instances/available). For inference at scale, RunPod's Serverless GPU Endpoints allow pay-per-token serving without pre-provisioned GPU capacity, a model unique among top-5 H100 providers. RunPod's H100 Secure Cloud pricing often undercuts Lambda by 15-25% for comparable configurations, but availability windows are shorter and provisioning can take 5-30 minutes during peak demand.

Pricing Analysis

RunPod's cheapest H100 option starts at $1.8900/hr, which is 48% below the market average of $3.6501/hr. RunPod offers both on-demand and spot pricing for H100. Spot instances can save 50–70% for interruptible workloads like batch processing and hyperparameter sweeps.

Note: RunPod's community cloud offers the lowest prices but with shared infrastructure — the secure cloud tier provides dedicated resources at moderately higher rates.

Market context: H100 is the most in-demand GPU in the cloud market as of 2026. Prices have stabilized after the 2024 shortage, but availability still varies significantly by provider and region.

Top Alternatives to RunPod for H100

Side-by-side cost comparison vs the cheapest RunPod H100 rate ($1.8900/hr).

Provider Price/hr Type vs RunPod
Vast.ai $1.4900/hr Spot -21% cheaper vs RunPod
Corvex $1.5900/hr Reserved -16% cheaper vs RunPod
DataCrunch $1.6500/hr Spot -13% cheaper vs RunPod
CoreWeave $1.7900/hr Reserved -5% cheaper vs RunPod
Shadeform $1.7900/hr On-Demand -5% cheaper vs RunPod

View all H100 providers →

Calculate Your Exact H100 Cost

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

● $19 — no subscription
Track RunPod H100 prices — daily alerts
Get emailed when RunPod drops below your budget. $19 for 30 days. No account needed.
Set Price Alert → Calculator →

Frequently Asked Questions

How much does RunPod H100 cost per hour in 2026? +
RunPod H100 starts at $1.8900/hr (spot). This is below the H100 market average of $3.6501/hr — RunPod is among the more competitively priced options for this GPU. RunPod offers 8 H100 configurations ranging from $1.8900 to $2.6900/hr. Use the GPU cost calculator to estimate total cost based on your hours and workload type.
What are the cheapest alternatives to RunPod for H100? +
There are 5 providers offering H100 cheaper than RunPod: Vast.ai at $1.4900/hr, Corvex at $1.5900/hr, DataCrunch at $1.6500/hr, and 2 more. 24 other providers offer H100 at higher rates. See the full comparison table for all H100 providers.
What pricing options does RunPod offer for H100? +
RunPod offers 2 pricing types for H100: spot, on-demand. Spot instances can save 50–70% on interruptible workloads like batch processing and hyperparameter sweeps. On-demand is better for production workloads that cannot be interrupted.
What workloads is RunPod H100 best suited for? +
The H100 (Hopper architecture, 80 GB HBM3) is best suited for Large language model training, foundation model pre-training, and high-throughput inference. RunPod is a marketplace provider — Individual developers, hobbyists, and cost-sensitive teams who prioritize price over enterprise features. RunPod's community cloud tier is often the cheapest option for experimentation.
How much does it cost to run H100 on RunPod for a month? +
At RunPod's cheapest H100 rate of $1.8900/hr: running 24/7 costs ~$45.36/day or ~$1379.70/month per GPU. For training jobs, costs scale linearly — a 100-hour training run would cost ~$189.00. 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 RunPod H100 — calculated from real 90-day pricing data, not an LLM.

Get Analysis — $19