ByteCosts

GPU rental pricing

RunPod GPU pricing: rental cost per hour by GPU and region

Last updated June 15, 2026 · ByteCosts

Direct answer

RunPod rents GPU instances from about $0.39 per GPU-hour on-demand (L4 in CA), or roughly $2.39 per GPU-hour on spot. This page tracks 688 instance rows across 10 GPU models (H100-NVL, L4, L40, A100-80GB, A100-80GB-SXM, B200, and more) and 8 regions, from the ByteCosts GPU catalog. Per-GPU rate is the instance hourly price divided by GPU count, so multi-GPU boxes can look cheaper per card. Spot and preemptible capacity is not guaranteed and can be reclaimed, so size on-demand for steady jobs and treat spot as a discount for interruptible work.

Compare GPU rental across providers - Cheapest per-GPU-hour by accelerator →

RunPod GPU rental at a glance

  • Cheapest on-demand: $0.39 per GPU-hour (L4, 10x_L4_SECURE, CA)
  • Cheapest spot: $2.39 per GPU-hour (H100)
  • 10 GPU models and 8 regions tracked
  • 688 instance rows, updated June 15, 2026

Cheapest RunPod rate by GPU model

Lowest on-demand per-GPU-hour rate for each GPU model RunPod offers, from the ByteCosts catalog. Per-GPU rate is the instance hourly price divided by GPU count.

RunPod cheapest on-demand rate per GPU model
GPU modelPer-GPU/hrInstanceGPUsRegionInstance/hr
L4$0.3910x_L4_SECURE10CA$3.90
L40$0.8210x_L40_SECURE10CA$8.20
L40S$0.861x_L40S_SECURE1CA$0.86
A100-80GB$1.391x_A100-80GB_SECURE1CA$1.39
A100-80GB-SXM$1.491x_A100-80GB-SXM_SECURE1CA$1.49
H100$2.891x_H100_SECURE1CA$2.89
H100-NVL$3.1910x_H100-NVL_SECURE10CA$31.90
H100-SXM$3.291x_H100-SXM_SECURE1CA$3.29
H200-SXM$4.391x_H200-SXM_SECURE1CA$4.39
B200$5.891x_B200_SECURE1CA$5.89

When RunPod GPUs are cost-effective

Headline per-hour price is only half the story for GPU rental. The variables that move a training or inference bill are GPU utilization (idle reserved capacity is pure waste), whether your job tolerates spot interruption, region and egress fees, and the storage and networking attached to the instance.

RunPod tends to win when you can keep the GPU busy, batch interruptible work onto spot, and pick a region close to your data. It tends to lose when capacity sits idle between jobs, when egress moves large datasets out repeatedly, or when a smaller or older GPU would clear your throughput target at a lower rate.

Hidden costs to watch with RunPod GPUs

The instance hourly rate rarely equals your invoice. Budget for these before committing:

  • Idle time: a reserved GPU bills whether or not a job is running, so low utilization multiplies effective cost per useful hour.
  • Spot reclaim: interruptible capacity can be taken back mid-job, forcing checkpoints and restarts that add wall-clock and waste.
  • Storage and networking: attached disks, snapshots, and inter-node bandwidth for multi-GPU training are billed on top of the GPU.
  • Egress: moving model weights or datasets out of the region adds per-GB transfer fees.
  • Region spread: the same GPU can cost noticeably more in scarce regions, so the cheapest row may not be where your data lives.

Frequently asked questions

How much does a RunPod GPU cost per hour?

RunPod rents GPUs from about $0.39 per GPU-hour on-demand for L4 (the 10x_L4_SECURE instance in CA), up to higher rates for newer accelerators. See the table above for the cheapest rate per GPU model. Rates are list rental prices before committed-use discounts.

Is spot or on-demand cheaper on RunPod?

Spot (interruptible) capacity on RunPod can drop to about $2.39 per GPU-hour, well below on-demand, but it can be reclaimed at any time. Use spot for checkpointed, restartable jobs and on-demand for steady serving where an interruption would be costly.

Where does ByteCosts get RunPod GPU prices?

Rates come from the SkyPilot open GPU catalog, which mirrors provider pricing, normalized into the ByteCosts GPU index and dated. The provider's own pricing API is the ground truth; the catalog is used for discovery and cross-checking, not re-export. Verify the provider source before production commitments.

Cite this page

RunPod GPU rental pricing. ByteCosts. Updated June 15, 2026. https://bytecosts.com/gpu/runpod/

Sources