GPU rental pricing
Modal GPU pricing: rental cost per hour by GPU and region
Modal rents GPU instances from about $0.59 per GPU-hour on-demand (T4 in published). This page tracks 11 instance rows across 11 GPU models (A10, A100, A100-80GB, B200, B300, H100-SXM, and more) and 1 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 →
Modal GPU rental at a glance
- Cheapest on-demand: $0.59 per GPU-hour (T4, nvidia-t4, published)
- Spot pricing not published for Modal in this catalog
- 11 GPU models and 1 regions tracked
- 11 instance rows, updated August 22, 2026
Cheapest Modal rate by GPU model
Lowest on-demand per-GPU-hour rate for each GPU model Modal offers, from the ByteCosts catalog. Per-GPU rate is the instance hourly price divided by GPU count.
| GPU model | Per-GPU/hr | Instance | GPUs | Region | Instance/hr |
|---|---|---|---|---|---|
| T4 | $0.59 | nvidia-t4 | 1 | published | $0.59 |
| L4 | $0.80 | nvidia-l4 | 1 | published | $0.80 |
| A10 | $1.10 | nvidia-a10 | 1 | published | $1.10 |
| L40S | $1.95 | nvidia-l40s | 1 | published | $1.95 |
| A100 | $2.10 | nvidia-a100-40-gb | 1 | published | $2.10 |
| A100-80GB | $2.50 | nvidia-a100-80-gb | 1 | published | $2.50 |
| RTXPRO6000 | $3.03 | nvidia-rtx-pro-6000 | 1 | published | $3.03 |
| H100-SXM | $3.95 | nvidia-h100-sxm5 | 1 | published | $3.95 |
| H200-SXM | $4.54 | nvidia-h200-sxm | 1 | published | $4.54 |
| B200 | $6.25 | nvidia-b200 | 1 | published | $6.25 |
| B300 | $7.10 | nvidia-b300 | 1 | published | $7.10 |
When Modal 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.
Modal 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.
Frequently asked questions
How much does a Modal GPU cost per hour?
Modal rents GPUs from about $0.59 per GPU-hour on-demand for T4 (the nvidia-t4 instance in published), 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 Modal?
Modal does not publish spot GPU rates in this catalog, so the on-demand rates above are the reference. Where a provider offers spot, it is cheaper but interruptible.
Where does ByteCosts get Modal GPU prices?
Rates come from the ByteCosts GPU catalog: SkyPilot’s open catalog plus first-party provider pricing pages and APIs, normalized to dollars per GPU-hour and dated. The provider’s own pricing page is the ground truth. Verify that source before production commitments.
Modal GPU rental pricing. ByteCosts. Updated August 22, 2026. https://bytecosts.com/gpu/modal/