GPU cloud comparison
GPU cloud pricing comparison
Compare 5476 instances across 466 GPU cloud providers and 137 accelerator types. Prices are shown per GPU instance hour where available.
| Provider | GPU | VRAM | Count | Instance | Price/hr | Tier | Region |
|---|---|---|---|---|---|---|---|
| Azure | A10 | 24GB | 1 | Standard_NV6ads_A10_v5 | $0.080 | spot | westus3 |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Spot | $0.080 | spot | eastus2 | |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Low Priority | $0.090 | spot | westus3 | |
| Vast.ai | RTX 3060 | 12GB | 1 | rtx3060.1x | $0.090 | community | us-central |
| Yotta Shakti | T4 | 16GB | 1 | yotta-shakti-t4-x1-eu-central-1-spot | $0.095 | spot | eu-central-1 |
| Cloudalize | RTX A4000 | 16GB | 1 | cloudalize-rtx-a4000-x1 | $0.099 | spot | US, Iowa |
| Vast.ai | RTX 5060 Ti | 16311 GB | 1 | — | $0.100 | on-demand | — |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Spot | $0.100 | spot | usgovvirginia | |
| Vast.ai | RTX A4000 | 16376 GB | 1 | — | $0.100 | on-demand | |
| Salad | RTX 3060 | 12GB | 1 | rtx3060-salad | $0.100 | community | global |
| CoreWeave Spark | T4 | 16GB | 1 | coreweave-spark-t4-x1-edge-community | $0.102 | community | edge |
| NeevCloud Reserved | T4 | 16GB | 1 | neevcloud-reserved-t4-x1-us-east-1-community | $0.106 | community | us-east-1 |
| Hyperstack Reserved | RTX 3090 | 24GB | 1 | hyperstack-reserved-rtx-3090-x1-eu-central-1-community | $0.109 | community | eu-central-1 |
| Foundry | T4 | 16GB | 1 | foundry-t4-x1 | $0.110 | spot | ca-central-1 |
| GCP | T4 | 16GB | 1 | n1 + T4 | $0.110 | spot | us-central1 |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Low Priority | $0.110 | spot | northcentralus | |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Low Priority | $0.110 | spot | usgovvirginia | |
| Vast.ai | RTX 4070 Ti | 12282 GB | 1 | — | $0.110 | on-demand | |
| Vast.ai | RTX 5060 Ti | 16311 GB | 1 | — | $0.110 | on-demand | |
| Vast.ai | RTX 3060 Ti | 8GB | 1 | rtx3060ti.1x | $0.110 | community | us-central |
| Lambda On-Demand Plus | Inferentia2 | 32GB | 1 | lambda-on-demand-plus-inferentia2-x1-ap-south-1-spot | $0.113 | spot | ap-south-1 |
| Phoenix GPU Cloud | Inferentia2 | 32GB | 1 | phoenix-gpu-cloud-inferentia2-x1-eu-west-2-spot | $0.118 | spot | eu-west-2 |
| Vast.ai | RTX 3080 Ti | 12288 GB | 1 | — | $0.120 | on-demand | — |
| Vast.ai | RTX 4070 Ti | 12282 GB | 1 | — | $0.120 | on-demand | — |
| Vast.ai | RTX 3090 | 24576 GB | 1 | — | $0.120 | on-demand | |
| RunPod | RTX A2000 | 6 GB | 1 | NVIDIA RTX A2000 | $0.120 | community | |
| RunPod | RTX A2000 | 6GB | 1 | NVIDIA RTX A2000 (community) | $0.120 | on-demand | community |
| RunPod | RTX A2000 | 6GB | 1 | NVIDIA RTX A2000 (community) | $0.120 | spot | community |
| Vast.ai | RTX 4060 | 8GB | 1 | rtx4060.1x | $0.120 | community | us-central |
| Salad | RTX 3060 Ti | 8GB | 1 | rtx3060ti-salad | $0.120 | community | global |
| Massed Compute | CPU | 1 | cpu_mini_amd_epyc | $0.120 | on-demand | US, Kansas City, KS | |
| Verda | CPU | 1 | CPU.4V.16G | $0.120 | on-demand | FI, Central | |
| Prefect Cloud GPU | RTX A4000 | 16GB | 1 | prefect-cloud-gpu-rtx-a4000-x1 | $0.121 | spot | eu-south-1 |
| Backblaze Compute | RTX 3090 | 24GB | 1 | backblaze-compute-rtx-3090-x1 | $0.125 | spot | eu-north-1 |
| OpalStack GPU | RTX A4000 | 16GB | 1 | opalstack-gpu-rtx-a4000-x1 | $0.128 | spot | US, California |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Spot | $0.130 | spot | westus3 | |
| RunPod | RTX 3070 | 8 GB | 1 | NVIDIA GeForce RTX 3070 | $0.130 | community | — |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Spot | $0.130 | spot | westus2 | |
| AWS | A100 | 8 | p4d.24xlarge | $0.130 | on-demand | us-east-1 | |
| AWS | H100 | 8 | p5.48xlarge | $0.130 | on-demand | us-east-1 | |
| RunPod | RTX 3070 | 8 GB | 1 | NVIDIA GeForce RTX 3070 | $0.130 | community | |
| RunPod | RTX 3070 | 8GB | 1 | NVIDIA GeForce RTX 3070 (community) | $0.130 | on-demand | community |
| RunPod | RTX 3070 | 8GB | 1 | NVIDIA GeForce RTX 3070 (community) | $0.130 | spot | community |
| RunPod | RTX 3060 | 12GB | 1 | rtx3060-community | $0.130 | community | us-east |
| Mara Cloud | RTX 3090 | 24GB | 1 | mara-cloud-rtx-3090-x1-af-south-1-community | $0.131 | community | af-south-1 |
| Foundry Reserved | RTX 4000 Ada | 20GB | 1 | foundry-reserved-rtx-4000-ada-x1-us-east-2-spot | $0.132 | spot | us-east-2 |
| Brev Reserved | RTX 4000 Ada | 20GB | 1 | brev-reserved-rtx-4000-ada-x1-us-west-1-spot | $0.136 | spot | us-west-1 |
| OpenSpire | RTX A5000 | 24GB | 1 | openspire-rtx-a5000-x1 | $0.136 | spot | US, Texas |
| SK Telecom Cloud | RTX 4080 | 16GB | 1 | sk-telecom-cloud-rtx-4080-x1 | $0.137 | spot | eu-central-1 |
| Vast.ai | RTX 4070S Ti | 16376 GB | 1 | — | $0.140 | on-demand | — |
| Vast.ai | RTX 5070 | 12227 GB | 1 | — | $0.140 | on-demand | — |
| Vast.ai | RTX 5070 Ti | 16303 GB | 1 | — | $0.140 | on-demand | |
| RunPod | RTX 3060 Ti | 8GB | 1 | rtx3060ti-community | $0.140 | community | us-east |
| Flowdesk Compute | RTX 4080 | 16GB | 1 | flowdesk-compute-rtx-4080-x1 | $0.141 | spot | EU, Stockholm |
| Featherless AI | Inferentia2 | 32GB | 1 | featherless-ai-inferentia2-x1-ap-northeast-1-community | $0.147 | community | ap-northeast-1 |
| Vast.ai | RTX 3090 | 24576 GB | 1 | — | $0.150 | on-demand | — |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Spot | $0.150 | spot | northcentralus | |
| Hyperstack | A4000 | 16GB | 1 | n3-RTX-A4000x1 | $0.150 | on-demand | NO, Oslo |
| TensorDock | RTX 3060 | 12GB | 1 | rtx3060-1x | $0.150 | spot | eu-west |
| Saturn Cloud | RTX A5000 | 24GB | 1 | saturn-cloud-rtx-a5000-x1 | $0.150 | spot | us-west-1 |
| Genesis Reserved | A40 | 48GB | 1 | genesis-reserved-a40-x1-global-spot | $0.150 | spot | global |
| Tenstorrent Cloud | RTX A5000 | 24GB | 1 | tenstorrent-cloud-rtx-a5000-x1-ap-northeast-1-spot | $0.151 | spot | ap-northeast-1 |
| Cake Reserved | Inferentia2 | 32GB | 1 | cake-reserved-inferentia2-x1-us-central-1-community | $0.153 | community | us-central-1 |
| Vultr Cloud GPU | RTX A4000 | 16GB | 1 | vultr-cloud-gpu-rtx-a4000-x1 | $0.154 | community | EU, Frankfurt |
| Tenstorrent Cloud | RTX 3090 | 24GB | 1 | tenstorrent-cloud-rtx-3090-x1-eu-west-2-reserved | $0.154 | reserved | eu-west-2 |
| Glaze AI | T4 | 16GB | 2 | glaze-ai-t4-x2-eu-central-1-spot | $0.157 | spot | eu-central-1 |
| AWS | T4 | 16GB | 1 | g4dn.xlarge | $0.158 | spot | us-east-1 |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Spot | $0.160 | spot | usgovarizona | |
| Vast.ai | RTX 5070 Ti | 16303 GB | 1 | — | $0.160 | on-demand | — |
| Azure | A10 | 1 | Standard_NV6ads_A10_v5 Spot | $0.160 | spot | westus | |
| Vast.ai | RTX 4060 Ti | 16380 GB | 2 | — | $0.160 | on-demand | |
| RunPod | RTX A5000 | 24 GB | 1 | NVIDIA RTX A5000 | $0.160 | community | |
| RunPod | RTX A5000 | 24GB | 1 | NVIDIA RTX A5000 (community) | $0.160 | on-demand | community |
| RunPod | RTX A5000 | 24GB | 1 | NVIDIA RTX A5000 (community) | $0.160 | spot | community |
| RunPod | RTX 4060 | 8GB | 1 | rtx4060-community | $0.160 | community | eu-central |
| Vast.ai | RTX 2080 Ti | 11GB | 1 | rtx2080ti.1x | $0.160 | community | us-central |
| Cerebrium Plus | RTX 4090 | 24GB | 1 | cerebrium-plus-rtx-4090-x1 | $0.164 | spot | ap-northeast-1 |
| Anyscale Reserved | L4 | 24GB | 1 | anyscale-reserved-l4-x1-ap-south-1-spot | $0.165 | spot | ap-south-1 |
| Hosteurope GPU | T4 | 16GB | 1 | hosteurope-gpu-t4-x1 | $0.167 | community | EU, Paris |
| Tata GPU-as-a-Service | T4 | 16GB | 1 | tata-gpu-as-a-service-t4-x1-ap-southeast-2-reserved | $0.168 | reserved | ap-southeast-2 |
| Covalent GPU | RTX 3080 | 10GB | 2 | covalent-gpu-rtx-3080-x2 | $0.169 | spot | EU, London |
| Azure | A10 | 1 | Standard_NV12ads_A10_v5 Spot | $0.170 | spot | eastus | |
| Azure | A10 | 1 | Standard_NV12ads_A10_v5 Spot | $0.170 | spot | westus3 | |
| RunPod | RTX 3080 | 10 GB | 1 | NVIDIA GeForce RTX 3080 | $0.170 | community | — |
| Azure | A10 | 1 | Standard_NV12ads_A10_v5 Spot | $0.170 | spot | eastus2 | |
| Azure | A10 | 1 | Standard_NV12ads_A10_v5 Spot | $0.170 | spot | westus2 | |
| RunPod | RTX 3080 | 10 GB | 1 | NVIDIA GeForce RTX 3080 | $0.170 | community | |
| RunPod | RTX A4000 | 16 GB | 1 | NVIDIA RTX A4000 | $0.170 | community | |
| RunPod | RTX 3080 | 10GB | 1 | NVIDIA GeForce RTX 3080 (community) | $0.170 | on-demand | community |
| RunPod | RTX 3080 | 10GB | 1 | NVIDIA GeForce RTX 3080 (community) | $0.170 | spot | community |
| RunPod | RTX A4000 | 16GB | 1 | NVIDIA RTX A4000 (community) | $0.170 | on-demand | community |
| RunPod | RTX A4000 | 16GB | 1 | NVIDIA RTX A4000 (community) | $0.170 | spot | community |
| SambaNova SambaCloud | T4 | 16GB | 1 | sambanova-sambacloud-t4-x1-us-west-1-reserved | $0.170 | reserved | us-west-1 |
| Innovecs Cloud | RTX 3090 | 24GB | 1 | innovecs-cloud-rtx-3090-x1 | $0.170 | community | EU, London |
| Hyperstack Reserved | Inferentia2 | 32GB | 1 | hyperstack-reserved-inferentia2-x1-eu-south-1-reserved-3y | $0.172 | reserved-3yr | eu-south-1 |
| SkyPilot Managed | A40 | 48GB | 1 | skypilot-managed-a40-x1-ap-southeast-1-spot | $0.174 | spot | ap-southeast-1 |
| Tenstorrent Cloud | T4 | 16GB | 1 | tenstorrent-cloud-t4-x1-ap-southeast-1-reserved | $0.174 | reserved | ap-southeast-1 |
| Hot Aisle Spark | RTX 3090 | 24GB | 1 | hot-aisle-spark-rtx-3090-x1-sa-east-1-reserved-1y | $0.176 | reserved-1yr | sa-east-1 |
| Brev Reserved | RTX A5000 | 24GB | 1 | brev-reserved-rtx-a5000-x1-ap-southeast-2-community | $0.176 | community | ap-southeast-2 |
| Positron AI Cloud | L4 | 24GB | 1 | positron-ai-cloud-l4-x1-us-east-2-spot | $0.178 | spot | us-east-2 |
| Yotta Shakti | RTX 4000 Ada | 20GB | 1 | yotta-shakti-rtx-4000-ada-x1-us-east-1-reserved-3y | $0.178 | reserved-3yr | us-east-1 |
| Vast.ai | RTX 4090 | 24GB | 1 | RTX 4090 | $0.180 | spot | marketplace |
| Vast.ai | RTX 3090 | 24GB | 1 | RTX 3090 | $0.180 | on-demand | marketplace |
| Vast.ai | RTX A4000 | 16GB | 1 | vast-rtxa4000 | $0.180 | community | global |
| RunPod | RTX 3080 Ti | 12 GB | 1 | NVIDIA GeForce RTX 3080 Ti | $0.180 | community | — |
| RunPod | RTX 4000 Ada SFF | 20 GB | 1 | NVIDIA RTX 4000 SFF Ada Generation | $0.180 | community | — |
| Azure | A10 | 1 | Standard_NV12ads_A10_v5 Low Priority | $0.180 | spot | eastus | |
| Azure | A10 | 1 | Standard_NV12ads_A10_v5 Low Priority | $0.180 | spot | westus2 | |
| Azure | A10 | 1 | Standard_NV12ads_A10_v5 Low Priority | $0.180 | spot | eastus2 | |
| Azure | A10 | 1 | Standard_NV12ads_A10_v5 Low Priority | $0.180 | spot | westus3 | |
| Salad | RTX 3090 | 24GB | 1 | RTX 3090 community | $0.180 | community | global |
| Vast.ai | RTX 4070S | 16376 GB | 1 | — | $0.180 | on-demand | |
| RunPod | RTX 3080 Ti | 12 GB | 1 | NVIDIA GeForce RTX 3080 Ti | $0.180 | community | |
| RunPod | RTX 4000 Ada SFF | 20 GB | 1 | NVIDIA RTX 4000 SFF Ada Generation | $0.180 | community | |
| RunPod | RTX 3080 Ti | 12GB | 1 | NVIDIA GeForce RTX 3080 Ti (community) | $0.180 | on-demand | community |
| RunPod | RTX 3080 Ti | 12GB | 1 | NVIDIA GeForce RTX 3080 Ti (community) | $0.180 | spot | community |
| RunPod | RTX 4000 Ada SFF | 20GB | 1 | NVIDIA RTX 4000 SFF Ada Generation (community) | $0.180 | on-demand | community |
| RunPod | RTX 4000 Ada SFF | 20GB | 1 | NVIDIA RTX 4000 SFF Ada Generation (community) | $0.180 | spot | community |
| Vast.ai | RTX 4070 | 12GB | 1 | rtx4070.1x | $0.180 | community | us-central |
| Tata GPU-as-a-Service | Inferentia2 | 32GB | 1 | tata-gpu-as-a-service-inferentia2-x1-eu-north-1-reserved-3y | $0.183 | reserved-3yr | eu-north-1 |
| RunPod | RTX 2000 Ada | 16GB | 1 | NVIDIA RTX 2000 Ada Generation (secure) | $0.184 | reserved-6mo | secure |
| Phoenix GPU Cloud | A10G | 24GB | 1 | phoenix-gpu-cloud-a10g-x1-eu-west-1-spot | $0.185 | spot | eu-west-1 |
| NetMagic | T4 | 16GB | 1 | netmagic-t4-x1 | $0.186 | community | sa-east-1 |
| Heroic Cloud | T4 | 16GB | 1 | heroic-cloud-t4 | $0.187 | community | na-east |
| RunPod | RTX 4070 Ti | 12 GB | 1 | NVIDIA GeForce RTX 4070 Ti | $0.190 | community | — |
| Vast.ai | RTX 4080 | 16376 GB | 1 | — | $0.190 | on-demand | |
| RunPod | RTX 4070 Ti | 12 GB | 1 | NVIDIA GeForce RTX 4070 Ti | $0.190 | community | |
| RunPod | RTX A4500 | 20 GB | 1 | NVIDIA RTX A4500 | $0.190 | community | |
| RunPod | Tesla V100 | 16 GB | 1 | Tesla V100-PCIE-16GB | $0.190 | community | |
| RunPod | RTX 4070 Ti | 12GB | 1 | NVIDIA GeForce RTX 4070 Ti (community) | $0.190 | on-demand | community |
| RunPod | RTX 4070 Ti | 12GB | 1 | NVIDIA GeForce RTX 4070 Ti (community) | $0.190 | spot | community |
| RunPod | RTX A4500 | 20GB | 1 | NVIDIA RTX A4500 (community) | $0.190 | on-demand | community |
| RunPod | RTX A4500 | 20GB | 1 | NVIDIA RTX A4500 (community) | $0.190 | spot | community |
| RunPod | Tesla V100 | 16GB | 1 | Tesla V100-PCIE-16GB (community) | $0.190 | on-demand | community |
| RunPod | Tesla V100 | 16GB | 1 | Tesla V100-PCIE-16GB (community) | $0.190 | spot | community |
| TensorDock | RTX 2080 Ti | 11GB | 1 | rtx2080ti-1x | $0.190 | spot | us-west |
| Latitude.sh | CPU | 1 | vm-small | $0.190 | on-demand | US, Dallas, TX | |
| Latitude.sh | CPU | 1 | vm-small | $0.190 | on-demand | TYO3 | |
| Latitude.sh | CPU | 1 | vm-small | $0.190 | on-demand | US, Ashburn, VA | |
| Latitude.sh | CPU | 1 | vm-small | $0.190 | on-demand | LON |
Last updated: · Source: aws pricing · azure pricing · gcp pricing · runpod pricing · vastai pricing · lambda pricing · coreweave pricing · nebius pricing · crusoe pricing · cudo pricing · oracle pricing · digitalocean pricing · voltagepark pricing · jarvislabs pricing · fly pricing · bitdeer pricing · denvr pricing · exabits pricing · taiga pricing · qarnot pricing · ori pricing · Methodology
Frequently asked questions
What do the GPU prices measure?
The recorded snapshot lists 5476 instances in USD per hour, grouped by provider, accelerator, GPU count, region and pricing tier. These are stored observations, not a guarantee that capacity is available; compare like-for-like instance configurations.
Are GPU prices quoted per token?
No. GPU rentals are billed per instance-hour. Normalize multi-GPU instances by GPU count for per-GPU-hour comparisons. Per-token cost also depends on throughput, utilization, model size, memory and operational overhead; do not compare an hourly rate directly with an API token rate.
How often is GPU pricing updated?
The GPU file records a weekly cadence and was last updated 2026-11-01. Rental-history observations are monthly. Opening this page does not verify or refresh provider quotes; the recorded snapshot date stays unchanged until data changes.
What additional GPU charges can apply?
Storage, data transfer, networking, taxes, minimum commitments and spot interruptions may affect the total bill. Reserved, spot and community tiers have different terms. Follow the linked provider sources and use your expected utilization before choosing a rental option.