Cloud GPU instances
Compare 351 cloud GPU configurations from 22 providers by GPU model, count, memory and fabric. 231 carry a sourced USD instance-hour rate with the GPU-hour worked out, and H100 capacity runs $2.50 to $12.29 per GPU-hour. Each instance page adds CPU, RAM, regions and the exact purchase terms behind the price.
Checked
| n3-RTX-A4000x1 | Hyperstack | NVIDIA RTX A4000 | 11 | 16 GB | $0.152 | $0.151 | Public USD GPU-hour list; NORWAY-1 preset | 2026-11-11 |
|---|---|---|---|---|---|---|---|---|
| RTX A5000 / 1 GPU Pod | RunPod | NVIDIA RTX A5000 | 11 | 24 GB | $0.271 | $0.271 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| n3-RTX-A4000x2 | Hyperstack | NVIDIA RTX A4000 | 21 | 16 GB | $0.302 | $0.151 | Public USD GPU-hour list; NORWAY-1 preset | 2026-11-11 |
| GPU+ (M4000) | Paperspace | NVIDIA M4000 | 11 | 8 GB | $0.452 | $0.451 | Paperspace public USD list | 2026-11-11 |
| n3-RTX-A6000x1 | Hyperstack | NVIDIA RTX A6000 | 11 | 48 GB | $0.502 | $0.501 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| RTX 3090 / 1 GPU Pod | RunPod | NVIDIA RTX 3090 | 11 | 24 GB | $0.501 | $0.501 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| P4000 | Paperspace | NVIDIA P4000 | 11 | 8 GB | $0.512 | $0.511 | Paperspace public USD list | 2026-11-11 |
| g4dn.xlarge | AWS | NVIDIA T4 | 11 | 16 GB | $0.5263 | $0.5261 | us-east-1 | 2026-11-11 |
| Standard_NC4as_T4_v3 | Azure | NVIDIA T4 | 11 | 16 GB | $0.5262 | $0.5261 | eastus | 2026-11-11 |
| RTX4000 | Paperspace | NVIDIA RTX4000 | 11 | 8 GB | $0.562 | $0.561 | Paperspace public USD list | 2026-11-11 |
| A40 / 1 GPU Pod | RunPod | NVIDIA A40 | 11 | 48 GB | $0.591 | $0.591 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| L4 / 1 GPU Pod | RunPod | NVIDIA L4 | 11 | 24 GB | $0.591 | $0.591 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| RTX A6000 / 1 GPU Pod | RunPod | NVIDIA RTX A6000 | 11 | 48 GB | $0.591 | $0.591 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| n3-RTX-A4000x4 | Hyperstack | NVIDIA RTX A4000 | 41 | 16 GB | $0.602 | $0.151 | Public USD GPU-hour list; NORWAY-1 preset | 2026-11-11 |
| rtx5000-28 | OVHcloud | NVIDIA Quadro-RTX5000 | 11 | 16 GB | $0.601 | $0.601 | Public USD list; deployment region not specified | 2026-11-11 |
| 1x Quadro RTX 6000 | Lambda | NVIDIA Quadro RTX 6000 | 11 | 24 GB | $0.691 | $0.691 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| g2-standard-4 | Google Cloud | NVIDIA L4 | 11 | 24 GB | $0.70682 | $0.70681 | us-central1 | 2026-11-11 |
| Standard_NC8as_T4_v3 | Azure | NVIDIA T4 | 11 | 16 GB | $0.7522 | $0.7521 | eastus | 2026-11-11 |
| A4000 | Paperspace | NVIDIA A4000 | 11 | 16 GB | $0.762 | $0.761 | Paperspace public USD list | 2026-11-11 |
| gpu-4000adax1-20gb | DigitalOcean | NVIDIA RTX 4000 Ada | 11 | 20 GB | $0.762 | $0.761 | Public USD on-demand list | 2026-11-11 |
| t1-45 | OVHcloud | NVIDIA Tesla V100 | 11 | 16 GB | $0.771 | $0.771 | Public USD list; deployment region not specified | 2026-11-11 |
| t1-le-45 | OVHcloud | NVIDIA Tesla V100 | 11 | 16 GB | $0.771 | $0.771 | Public USD list; deployment region not specified | 2026-11-11 |
| P5000 | Paperspace | NVIDIA P5000 | 11 | 16 GB | $0.782 | $0.781 | Paperspace public USD list | 2026-11-11 |
| g6.xlarge | AWS | NVIDIA L4 | 11 | 24 GB | $0.80483 | $0.80481 | us-east-1 | 2026-11-11 |
| L40 / 1 GPU Pod | RunPod | NVIDIA L40 | 11 | 48 GB | $0.821 | $0.821 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| RTX5000 | Paperspace | NVIDIA RTX5000 | 11 | 16 GB | $0.822 | $0.821 | Paperspace public USD list | 2026-11-11 |
| g2-standard-8 | Google Cloud | NVIDIA L4 | 11 | 24 GB | $0.85362 | $0.85361 | us-central1 (Iowa) | 2026-11-11 |
| t2-45 | OVHcloud | NVIDIA Tesla V100S | 11 | 32 GB | $0.881 | $0.881 | Public USD list; deployment region not specified | 2026-11-11 |
| t2-le-45 | OVHcloud | NVIDIA Tesla V100S | 11 | 32 GB | $0.881 | $0.881 | Public USD list; deployment region not specified | 2026-11-11 |
| RTX 4090 / 1 GPU Pod | RunPod | NVIDIA RTX 4090 | 11 | 24 GB | $0.891 | $0.891 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| g6.2xlarge | AWS | NVIDIA L4 | 11 | 24 GB | $0.97763 | $0.97761 | us-east-1 | 2026-11-11 |
| RTX 6000 Ada / 1 GPU Pod | RunPod | NVIDIA RTX 6000 Ada | 11 | 48 GB | $0.991 | $0.991 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| a10-45 | OVHcloud | NVIDIA A10 | 11 | 24 GB | $1.001 | $1.001 | Public USD list; deployment region not specified | 2026-11-11 |
| l4-90 | OVHcloud | NVIDIA L4 | 11 | 24 GB | $1.001 | $1.001 | Public USD list; deployment region not specified | 2026-11-11 |
| n3-L40x1 | Hyperstack | NVIDIA L40 | 11 | 48 GB | $1.002 | $1.001 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| n3-RTX-A6000x2 | Hyperstack | NVIDIA RTX A6000 | 21 | 48 GB | $1.002 | $0.501 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| g2-standard-12 | Google Cloud | NVIDIA L4 | 11 | 24 GB | $1.00042 | $1.00041 | us-central1 (Iowa) | 2026-11-11 |
| g5.xlarge | AWS | NVIDIA A10G | 11 | 24 GB | $1.0063 | $1.0061 | us-east-1 | 2026-11-11 |
| 1x A6000 | Lambda | NVIDIA A6000 | 11 | 48 GB | $1.091 | $1.091 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| L40S / 1 GPU Pod | RunPod | NVIDIA L40S | 11 | 48 GB | $1.091 | $1.091 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| P6000 | Paperspace | NVIDIA P6000 | 11 | 24 GB | $1.102 | $1.101 | Paperspace public USD list | 2026-11-11 |
| g2-standard-16 | Google Cloud | NVIDIA L4 | 11 | 24 GB | $1.14722 | $1.14721 | us-central1 (Iowa) | 2026-11-11 |
| RTX 5090 / 1 GPU Pod | RunPod | NVIDIA RTX 5090 | 11 | 32 GB | $1.191 | $1.191 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| n3-RTX-A4000x8 | Hyperstack | NVIDIA RTX A4000 | 81 | 16 GB | $1.202 | $0.151 | Public USD GPU-hour list; NORWAY-1 preset | 2026-11-11 |
| rtx5000-56 | OVHcloud | NVIDIA Quadro-RTX5000 | 21 | 16 GB | $1.201 | $0.601 | Public USD list; deployment region not specified | 2026-11-11 |
| Standard_NC16as_T4_v3 | Azure | NVIDIA T4 | 11 | 16 GB | $1.2042 | $1.2041 | eastus | 2026-11-11 |
| 1x A10 | Lambda | NVIDIA A10 | 11 | 24 GB | $1.291 | $1.291 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| g6.4xlarge | AWS | NVIDIA L4 | 11 | 24 GB | $1.32323 | $1.32321 | us-east-1 | 2026-11-11 |
| n3-A100x1 | Hyperstack | NVIDIA A100 PCIe | 11 | 80 GB | $1.352 | $1.351 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| A5000 | Paperspace | NVIDIA A5000 | 11 | 24 GB | $1.382 | $1.381 | Paperspace public USD list | 2026-11-11 |
| l40s-48gb.1x | Crusoe | NVIDIA L40S | 11 | 48 GB | $1.503 | $1.501 | Crusoe public USD GPU-hour list | 2026-11-11 |
| n3-RTX-A4000x10 | Hyperstack | NVIDIA RTX A4000 | 101 | 16 GB | $1.502 | $0.151 | Public USD GPU-hour list; NORWAY-1 preset | 2026-11-11 |
| t1-90 | OVHcloud | NVIDIA Tesla V100 | 21 | 16 GB | $1.551 | $0.7751 | Public USD list; deployment region not specified | 2026-11-11 |
| t1-le-90 | OVHcloud | NVIDIA Tesla V100 | 21 | 16 GB | $1.551 | $0.7751 | Public USD list; deployment region not specified | 2026-11-11 |
| gpu-6000adax1-48gb | DigitalOcean | NVIDIA RTX 6000 Ada | 11 | 48 GB | $1.572 | $1.571 | Public USD on-demand list | 2026-11-11 |
| gpu-l40sx1-48gb | DigitalOcean | NVIDIA L40S | 11 | 48 GB | $1.572 | $1.571 | Public USD on-demand list | 2026-11-11 |
| vcg-l40s-16c-180g-48vram | Vultr | NVIDIA L40S | 11 | 48 GB | $1.6711 | $1.6711 | Public API catalogue; locations array empty | 2026-11-11 |
| g2-standard-32 | Google Cloud | NVIDIA L4 | 11 | 24 GB | $1.73442 | $1.73441 | us-central1 (Iowa) | 2026-11-11 |
| t2-90 | OVHcloud | NVIDIA Tesla V100S | 21 | 32 GB | $1.761 | $0.881 | Public USD list; deployment region not specified | 2026-11-11 |
| t2-le-90 | OVHcloud | NVIDIA Tesla V100S | 21 | 32 GB | $1.761 | $0.881 | Public USD list; deployment region not specified | 2026-11-11 |
| A100 PCIe / 1 GPU Pod | RunPod | NVIDIA A100 PCIe | 11 | 80 GB | $1.791 | $1.791 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| A100 SXM / 1 GPU Pod | RunPod | NVIDIA A100 SXM | 11 | 80 GB | $1.791 | $1.791 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| l40s-90 | OVHcloud | NVIDIA L40S | 11 | 48 GB | $1.801 | $1.801 | Public USD list; deployment region not specified | 2026-11-11 |
| rtx5000-84 | OVHcloud | NVIDIA Quadro-RTX5000 | 31 | 16 GB | $1.801 | $0.601 | Public USD list; deployment region not specified | 2026-11-11 |
| n3-RTX-PRO6000-SEx1 | Hyperstack | NVIDIA RTX PRO 6000 SE | 11 | 96 GB | $1.852 | $1.851 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| g6e.xlarge | AWS | NVIDIA L40S | 11 | 48 GB | $1.8613 | $1.8611 | us-east-1 | 2026-11-11 |
| A6000 | Paperspace | NVIDIA A6000 | 11 | 48 GB | $1.892 | $1.891 | Paperspace public USD list | 2026-11-11 |
| 1x A100 PCIe | Lambda | NVIDIA A100 PCIe | 11 | 40 GB | $1.991 | $1.991 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| 1x A100 SXM | Lambda | NVIDIA A100 SXM | 11 | 40 GB | $1.991 | $1.991 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| a10-90 | OVHcloud | NVIDIA A10 | 21 | 24 GB | $2.001 | $1.001 | Public USD list; deployment region not specified | 2026-11-11 |
| a100-80gb.1x | Crusoe | NVIDIA A100 | 11 | 80 GB | $2.003 | $2.001 | Crusoe public USD GPU-hour list | 2026-11-11 |
| l4-180 | OVHcloud | NVIDIA L4 | 21 | 24 GB | $2.001 | $1.001 | Public USD list; deployment region not specified | 2026-11-11 |
| n3-L40x2 | Hyperstack | NVIDIA L40 | 21 | 48 GB | $2.002 | $1.001 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| n3-RTX-A6000x4 | Hyperstack | NVIDIA RTX A6000 | 41 | 48 GB | $2.002 | $0.501 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| VM.GPU.A10.1 | Oracle | NVIDIA A10 | 11 | 24 GB | $2.002 | $2.001 | OCI USD list | 2026-11-11 |
| g2-standard-24 | Google Cloud | NVIDIA L4 | 21 | 24 GB | $2.00082 | $1.00041 | us-central1 | 2026-11-11 |
| g6.8xlarge | AWS | NVIDIA L4 | 11 | 24 GB | $2.01443 | $2.01441 | us-east-1 | 2026-11-11 |
| 2x A6000 | Lambda | NVIDIA A6000 | 21 | 48 GB | $2.181 | $1.091 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| g6e.2xlarge | AWS | NVIDIA L40S | 11 | 48 GB | $2.24213 | $2.24211 | us-east-1 | 2026-11-11 |
| V100 | Paperspace | NVIDIA V100 | 11 | 16 GB | $2.302 | $2.301 | Paperspace public USD list | 2026-11-11 |
| V100-32G | Paperspace | NVIDIA V100 | 11 | 32 GB | $2.302 | $2.301 | Paperspace public USD list | 2026-11-11 |
| RTX Pro 6000 / 1 GPU Pod | RunPod | NVIDIA RTX Pro 6000 | 11 | 96 GB | $2.491 | $2.491 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| n3-H100x1 | Hyperstack | NVIDIA H100 PCIe | 11 | 80 GB | $2.502 | $2.501 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| 1x GH200 | Lambda | NVIDIA GH200 | 11 | 96 GB | $2.591 | $2.591 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| gpu-mi300x1-192gb | DigitalOcean | AMD MI300X | 11 | 192 GB | $2.592 | $2.591 | Public USD on-demand list | 2026-11-11 |
| n3-A100x2 | Hyperstack | NVIDIA A100 PCIe | 21 | 80 GB | $2.702 | $1.351 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| H100 PCIe / 1 GPU Pod | RunPod | NVIDIA H100 PCIe | 11 | 80 GB | $2.891 | $2.891 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| h100-380 | OVHcloud | NVIDIA H100 | 11 | 80 GB | $2.991 | $2.991 | Public USD list; deployment region not specified | 2026-11-11 |
| l40s-48gb.2x | Crusoe | NVIDIA L40S | 21 | 48 GB | $3.003 | $1.501 | Crusoe public USD GPU-hour list | 2026-11-11 |
| g6e.4xlarge | AWS | NVIDIA L40S | 11 | 48 GB | $3.00423 | $3.00421 | us-east-1 | 2026-11-11 |
| a100-180 | OVHcloud | NVIDIA A100 | 11 | 80 GB | $3.071 | $3.071 | Public USD list; deployment region not specified | 2026-11-11 |
| A100 | Paperspace | NVIDIA A100 | 11 | 40 GB | $3.092 | $3.091 | Paperspace public USD list | 2026-11-11 |
| t1-180 | OVHcloud | NVIDIA Tesla V100 | 41 | 16 GB | $3.101 | $0.7751 | Public USD list; deployment region not specified | 2026-11-11 |
| t1-le-180 | OVHcloud | NVIDIA Tesla V100 | 41 | 16 GB | $3.101 | $0.7751 | Public USD list; deployment region not specified | 2026-11-11 |
| A100-80G | Paperspace | NVIDIA A100 | 11 | 80 GB | $3.182 | $3.181 | Paperspace public USD list | 2026-11-11 |
| H100 NVL / 1 GPU Pod | RunPod | NVIDIA H100 NVL | 11 | 94 GB | $3.191 | $3.191 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| g7e.2xlarge | AWS | NVIDIA RTX PRO 6000 | 11 | 96 GB | $3.36313 | $3.36311 | us-east-1 | 2026-11-11 |
| g6.16xlarge | AWS | NVIDIA L4 | 11 | 24 GB | $3.39683 | $3.39681 | us-east-1 | 2026-11-11 |
| t2-180 | OVHcloud | NVIDIA Tesla V100S | 41 | 32 GB | $3.531 | $0.88251 | Public USD list; deployment region not specified | 2026-11-11 |
| t2-le-180 | OVHcloud | NVIDIA Tesla V100S | 41 | 32 GB | $3.531 | $0.88251 | Public USD list; deployment region not specified | 2026-11-11 |
| l40s-180 | OVHcloud | NVIDIA L40S | 21 | 48 GB | $3.601 | $1.801 | Public USD list; deployment region not specified | 2026-11-11 |
| Standard_NC24ads_A100_v4 | Azure | NVIDIA A100 PCIe | 11 | 80 GB | $3.6732 | $3.6731 | eastus | 2026-11-11 |
| a2-highgpu-1g | Google Cloud | NVIDIA A100 | 11 | 40 GB | $3.67342 | $3.67341 | us-central1 | 2026-11-11 |
| n3-RTX-PRO6000-SEx2 | Hyperstack | NVIDIA RTX PRO 6000 SE | 21 | 96 GB | $3.702 | $1.851 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| 1x H100 PCIe | Lambda | NVIDIA H100 PCIe | 11 | 80 GB | $3.791 | $3.791 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| MI325X / 1 GPU Droplet | DigitalOcean | AMD MI325X | 11 | 256 GB | $3.801 | $3.801 | Public USD on-demand GPU Droplets list | 2026-11-11 |
| 2x A100 PCIe | Lambda | NVIDIA A100 PCIe | 21 | 40 GB | $3.981 | $1.991 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| H100 SXM / 1 GPU Pod | RunPod | NVIDIA H100 SXM | 11 | 80 GB | $3.991 | $3.991 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| a10-180 | OVHcloud | NVIDIA A10 | 41 | 24 GB | $4.001 | $1.001 | Public USD list; deployment region not specified | 2026-11-11 |
| a100-80gb.2x | Crusoe | NVIDIA A100 | 21 | 80 GB | $4.003 | $2.001 | Crusoe public USD GPU-hour list | 2026-11-11 |
| l4-360 | OVHcloud | NVIDIA L4 | 41 | 24 GB | $4.001 | $1.001 | Public USD list; deployment region not specified | 2026-11-11 |
| n3-L40x4 | Hyperstack | NVIDIA L40 | 41 | 48 GB | $4.002 | $1.001 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| n3-RTX-A6000x8 | Hyperstack | NVIDIA RTX A6000 | 81 | 48 GB | $4.002 | $0.501 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| VM.GPU.A10.2 | Oracle | NVIDIA A10 | 21 | 24 GB | $4.002 | $2.002 | OCI USD list | 2026-11-11 |
| g2-standard-48 | Google Cloud | NVIDIA L4 | 41 | 24 GB | $4.00172 | $1.00041 | us-central1 | 2026-11-11 |
| Standard_NC64as_T4_v3 | Azure | NVIDIA T4 | 41 | 16 GB | $4.3522 | $1.0881 | eastus | 2026-11-11 |
| 4x A6000 | Lambda | NVIDIA A6000 | 41 | 48 GB | $4.361 | $1.091 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| gpu-h100x1-80gb | DigitalOcean | NVIDIA H100 | 11 | 80 GB | $4.412 | $4.411 | Public USD on-demand list | 2026-11-11 |
| gpu-h200x1-141gb | DigitalOcean | NVIDIA H200 | 11 | 141 GB | $4.472 | $4.471 | Public USD on-demand list | 2026-11-11 |
| gpu-h100-sxm / 1gpu-16vcpu-200gb | Nebius | NVIDIA H100 | 11 | 80 GB | $4.502 | $4.501 | eu-north1 | 2026-11-11 |
| g6e.8xlarge | AWS | NVIDIA L40S | 11 | 48 GB | $4.52863 | $4.52861 | us-east-1 | 2026-11-11 |
| g6.12xlarge | AWS | NVIDIA L4 | 41 | 24 GB | $4.60163 | $1.15041 | us-east-1 | 2026-11-11 |
| 1x H100 SXM | Lambda | NVIDIA H100 SXM | 11 | 80 GB | $4.891 | $4.891 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| n3-H100x2 | Hyperstack | NVIDIA H100 PCIe | 21 | 80 GB | $5.002 | $2.501 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| a2-ultragpu-1g | Google Cloud | NVIDIA A100 | 11 | 80 GB | $5.06882 | $5.06881 | us-central1 | 2026-11-11 |
| H200 / 1 GPU Pod | RunPod | NVIDIA H200 | 11 | 141 GB | $5.291 | $5.291 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| gpu-h200-sxm / 1gpu-16vcpu-200gb | Nebius | NVIDIA H200 | 11 | 141 GB | $5.402 | $5.401 | eu-north1 | 2026-11-11 |
| n3-A100x4 | Hyperstack | NVIDIA A100 PCIe | 41 | 80 GB | $5.402 | $1.351 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| g5.12xlarge | AWS | NVIDIA A10G | 41 | 24 GB | $5.6723 | $1.4181 | us-east-1 | 2026-11-11 |
| H100 | Paperspace | NVIDIA H100 | 11 | 80 GB | $5.952 | $5.951 | Paperspace public USD list | 2026-11-11 |
| h100-760 | OVHcloud | NVIDIA H100 | 21 | 80 GB | $5.981 | $2.991 | Public USD list; deployment region not specified | 2026-11-11 |
| l40s-48gb.4x | Crusoe | NVIDIA L40S | 41 | 48 GB | $6.003 | $1.501 | Crusoe public USD GPU-hour list | 2026-11-11 |
| a100-360 | OVHcloud | NVIDIA A100 | 21 | 80 GB | $6.151 | $3.0751 | Public USD list; deployment region not specified | 2026-11-11 |
| 8x Tesla V100 | Lambda | NVIDIA Tesla V100 | 81 | 16 GB | $6.321 | $0.791 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| g6.24xlarge | AWS | NVIDIA L4 | 41 | 24 GB | $6.67523 | $1.66881 | us-east-1 | 2026-11-11 |
| p5.4xlarge | AWS | NVIDIA H100 | 11 | 80 GB | $6.883 | $6.881 | us-east-1 | 2026-11-11 |
| Standard_NC40ads_H100_v5 | Azure | NVIDIA H100 NVL | 11 | 94 GB | $6.982 | $6.981 | eastus | 2026-11-11 |
| l40s-360 | OVHcloud | NVIDIA L40S | 41 | 48 GB | $7.201 | $1.801 | Public USD list; deployment region not specified | 2026-11-11 |
| Standard_NC48ads_A100_v4 | Azure | NVIDIA A100 PCIe | 21 | 80 GB | $7.3462 | $3.6731 | eastus | 2026-11-11 |
| n3-RTX-PRO6000-SEx4 | Hyperstack | NVIDIA RTX PRO 6000 SE | 41 | 96 GB | $7.402 | $1.851 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| g6e.16xlarge | AWS | NVIDIA L40S | 11 | 48 GB | $7.57723 | $7.57721 | us-east-1 | 2026-11-11 |
| 4x A100 PCIe | Lambda | NVIDIA A100 PCIe | 41 | 40 GB | $7.961 | $1.991 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| 1x B200 SXM6 | Lambda | NVIDIA B200 SXM6 | 11 | 180 GB | $7.991 | $7.991 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| B200 / 1 GPU Pod | RunPod | NVIDIA B200 | 11 | 180 GB | $7.991 | $7.991 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| a100-80gb.4x | Crusoe | NVIDIA A100 | 41 | 80 GB | $8.003 | $2.001 | Crusoe public USD GPU-hour list | 2026-11-11 |
| BM.GPU.A10.4 | Oracle | NVIDIA A10 | 41 | 24 GB | $8.002 | $2.002 | OCI USD list | 2026-11-11 |
| n3-L40x8 | Hyperstack | NVIDIA L40 | 81 | 48 GB | $8.002 | $1.001 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| g2-standard-96 | Google Cloud | NVIDIA L4 | 81 | 24 GB | $8.00332 | $1.00041 | us-central1 | 2026-11-11 |
| gpu-b200-sxm / 1gpu-20vcpu-224gb | Nebius | NVIDIA B200 | 11 | 180 GB | $8.502 | $8.501 | us-central1 | 2026-11-11 |
| B300 / 1 GPU Pod | RunPod | NVIDIA B300 | 11 | 288 GB | $8.991 | $8.991 | Secure Cloud public USD Pod list; region-specific stock | 2026-11-11 |
| 2x H100 SXM | Lambda | NVIDIA H100 SXM | 21 | 80 GB | $9.581 | $4.791 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| n3-H100x4 | Hyperstack | NVIDIA H100 PCIe | 41 | 80 GB | $10.002 | $2.501 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| g6e.12xlarge | AWS | NVIDIA L40S | 41 | 48 GB | $10.49263 | $2.62321 | us-east-1 | 2026-11-11 |
| n3-A100x8 | Hyperstack | NVIDIA A100 PCIe | 81 | 80 GB | $10.802 | $1.351 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| n3-A100x8-NVLink | Hyperstack | NVIDIA A100 NVLink | 81 | 80 GB | $11.202 | $1.401 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| h100-1520 | OVHcloud | NVIDIA H100 | 41 | 80 GB | $11.971 | $2.99251 | Public USD list; deployment region not specified | 2026-11-11 |
| l40s-48gb.8x | Crusoe | NVIDIA L40S | 81 | 48 GB | $12.003 | $1.501 | Crusoe public USD GPU-hour list | 2026-11-11 |
| a100-720 | OVHcloud | NVIDIA A100 | 41 | 80 GB | $12.291 | $3.07251 | Public USD list; deployment region not specified | 2026-11-11 |
| n3-A100-SXM4x8 | Hyperstack | NVIDIA A100 SXM | 81 | 80 GB | $12.802 | $1.601 | Public USD GPU-hour list; US-1 preset | 2026-11-11 |
| g6.48xlarge | AWS | NVIDIA L4 | 81 | 24 GB | $13.35043 | $1.66881 | us-east-1 | 2026-11-11 |
| Standard_NC80adis_H100_v5 | Azure | NVIDIA H100 NVL | 21 | 94 GB | $13.962 | $6.981 | eastus | 2026-11-11 |
| BM.GPU.L40S.4 | Oracle | NVIDIA L40S | 41 | 48 GB | $14.002 | $3.502 | OCI USD list | 2026-11-11 |
| Standard_NC96ads_A100_v4 | Azure | NVIDIA A100 PCIe | 41 | 80 GB | $14.6922 | $3.6731 | eastus | 2026-11-11 |
| n3-RTX-PRO6000-SEx8 | Hyperstack | NVIDIA RTX PRO 6000 SE | 81 | 96 GB | $14.802 | $1.851 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| l40s-48gb.10x | Crusoe | NVIDIA L40S | 101 | 48 GB | $15.003 | $1.501 | Crusoe public USD GPU-hour list | 2026-11-11 |
| g6e.24xlarge | AWS | NVIDIA L40S | 41 | 48 GB | $15.06563 | $3.76641 | us-east-1 | 2026-11-11 |
| 2x B200 SXM6 | Lambda | NVIDIA B200 SXM6 | 21 | 180 GB | $15.781 | $7.891 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| 8x HGX A100 40GB | Lambda | NVIDIA HGX A100 | 81 | 40 GB | $15.921 | $1.991 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| a100-80gb.8x | Crusoe | NVIDIA A100 | 81 | 80 GB | $16.003 | $2.001 | Crusoe public USD GPU-hour list | 2026-11-11 |
| g5.48xlarge | AWS | NVIDIA A10G | 81 | 24 GB | $16.2883 | $2.0361 | us-east-1 | 2026-11-11 |
| gd-8xl40s-i128 | CoreWeave | NVIDIA L40S | 81 | 48 GB | $18.002 | $2.251 | North America | 2026-11-11 |
| a100-80gb-sxm-ib.8x | Crusoe | NVIDIA A100 | 81 | 80 GB | $18.404 | $2.301 | Crusoe public USD GPU-hour list | 2026-11-11 |
| 4x H100 SXM | Lambda | NVIDIA H100 SXM | 41 | 80 GB | $18.761 | $4.691 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| n3-H100x8 | Hyperstack | NVIDIA H100 PCIe | 81 | 80 GB | $20.002 | $2.501 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| gpu-mi300x8-1536gb | DigitalOcean | AMD MI300X | 81 | 192 GB | $20.722 | $2.591 | Public USD on-demand list | 2026-11-11 |
| n3-H100x8-NVLink | Hyperstack | NVIDIA H100 NVLink | 81 | 80 GB | $20.802 | $2.601 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| p4d.24xlarge | AWS | NVIDIA A100 | 81 | 40 GB | $21.95763 | $2.74471 | us-east-1 | 2026-11-11 |
| BM.GPU4.8 | Oracle | NVIDIA A100 | 81 | 40 GB | $24.402 | $3.051 | OCI USD list | 2026-11-11 |
| 8x HGX A100 80GB | Lambda | NVIDIA HGX A100 | 81 | 80 GB | $25.521 | $3.191 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| n3-H100-SXM5x8 | Hyperstack | NVIDIA H100 SXM | 81 | 80 GB | $25.602 | $3.201 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| p4de.24xlarge | AWS | NVIDIA A100 | 81 | 80 GB | $27.44713 | $3.43091 | us-east-1 | 2026-11-11 |
| mi300x-192gb-ib.8x | Crusoe | AMD MI300X | 81 | 192 GB | $27.604 | $3.451 | Crusoe public USD GPU-hour list | 2026-11-11 |
| a2-highgpu-8g | Google Cloud | NVIDIA A100 | 81 | 40 GB | $29.38712 | $3.67341 | us-central1 | 2026-11-11 |
| g6e.48xlarge | AWS | NVIDIA L40S | 81 | 48 GB | $30.13123 | $3.76641 | us-east-1 | 2026-11-11 |
| MI325X / 8 GPU Droplet | DigitalOcean | AMD MI325X | 81 | 256 GB | $30.401 | $3.801 | Public USD on-demand GPU Droplets list | 2026-11-11 |
| 4x B200 SXM6 | Lambda | NVIDIA B200 SXM6 | 41 | 180 GB | $31.161 | $7.791 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| h100-80gb-sxm-ib.8x | Crusoe | NVIDIA H100 SXM | 81 | 80 GB | $31.205 | $3.901 | Crusoe public USD GPU-hour list | 2026-11-11 |
| HGX H100 SXM / 8 GPU node | Together | NVIDIA H100 SXM | 82 | 80 GB | $31.923 | $3.992 | Public USD GPU Clusters list | 2026-11-11 |
| n3-H200-SXM5x8 | Hyperstack | NVIDIA H200 SXM | 81 | 141 GB | $31.922 | $3.991 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| BM.GPU.A100-v2.8 | Oracle | NVIDIA A100 | 81 | 80 GB | $32.002 | $4.001 | OCI USD list | 2026-11-11 |
| Standard_ND96amsr_A100_v4 | Azure | NVIDIA A100 | 81 | 80 GB | $32.772 | $4.09631 | eastus | 2026-11-11 |
| h200-141gb-sxm-ib.8x | Crusoe | NVIDIA H200 | 81 | 141 GB | $34.324 | $4.291 | Crusoe public USD GPU-hour list | 2026-11-11 |
| gpu-h100x8-640gb | DigitalOcean | NVIDIA H100 | 81 | 80 GB | $35.282 | $4.411 | Public USD on-demand list | 2026-11-11 |
| gpu-h200x8-1128gb | DigitalOcean | NVIDIA H200 | 81 | 141 GB | $35.762 | $4.471 | Public USD on-demand list | 2026-11-11 |
| BM.GPU.RTXPRO.8 | Oracle | NVIDIA RTX PRO 6000 | 81 | 96 GB | $36.002 | $4.502 | OCI USD list | 2026-11-11 |
| gpu-h100-sxm / 8gpu-128vcpu-1600gb | Nebius | NVIDIA H100 | 81 | 80 GB | $36.002 | $4.501 | eu-north1 | 2026-11-11 |
| 8x HGX H100 | Lambda | NVIDIA HGX H100 | 81 | 80 GB | $36.721 | $4.591 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| a2-ultragpu-8g | Google Cloud | NVIDIA A100 | 81 | 80 GB | $40.55042 | $5.06881 | us-central1 | 2026-11-11 |
| gpu-h200-sxm / 8gpu-128vcpu-1600gb | Nebius | NVIDIA H200 | 81 | 141 GB | $43.202 | $5.401 | eu-north1 | 2026-11-11 |
| H100x8 | Paperspace | NVIDIA H100 | 81 | 80 GB | $47.602 | $5.951 | Paperspace public USD list | 2026-11-11 |
| HGX H200 / 8 GPU node | Together | NVIDIA H200 | 82 | 141 GB | $47.923 | $5.992 | Public USD GPU Clusters list | 2026-11-11 |
| BM.GPU.MI300X.8 | Oracle | AMD MI300X | 82 | 192 GB | $48.003 | $6.003 | OCI USD list | 2026-11-11 |
| n3-B200-SXM6x8 | Hyperstack | NVIDIA B200 | 81 | 192 GB | $48.002 | $6.001 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| gd-8xh100ib-i128 | CoreWeave | NVIDIA H100 | 81 | 80 GB | $49.242 | $6.1551 | North America | 2026-11-11 |
| h200-1920 | OVHcloud | NVIDIA H200 | 81 | 141 GB | $49.561 | $6.1951 | Public USD list; deployment region not specified | 2026-11-11 |
| h200-1920-eph | OVHcloud | NVIDIA H200 | 81 | 141 GB | $49.561 | $6.1951 | Public USD list; deployment region not specified | 2026-11-11 |
| gd-8xh200ib-i128 | CoreWeave | NVIDIA H200 | 81 | 141 GB | $50.442 | $6.3051 | North America | 2026-11-11 |
| p5.48xlarge | AWS | NVIDIA H100 | 81 | 80 GB | $55.043 | $6.881 | us-east-1 | 2026-11-11 |
| a2-megagpu-16g | Google Cloud | NVIDIA A100 | 161 | 40 GB | $55.73952 | $3.48371 | us-central1 | 2026-11-11 |
| n3-B300-SXM6x8 | Hyperstack | NVIDIA B300 | 81 | 288 GB | $59.202 | $7.401 | Public USD GPU-hour list; CANADA-1 preset | 2026-11-11 |
| 8x HGX B200 | Lambda | NVIDIA HGX B200 | 81 | 180 GB | $61.521 | $7.691 | Public USD list, effective 16 Nov 2026 | 2026-11-11 |
| p5en.48xlarge | AWS | NVIDIA H200 | 81 | 141 GB | $63.2963 | $7.9121 | us-east-1 | 2026-11-11 |
| BM.GPU.GB200.4 | Oracle | NVIDIA GB200 | 41 | 192 GB | $64.002 | $16.002 | OCI USD list | 2026-11-11 |
| HGX B200 / 8 GPU node | Together | NVIDIA B200 | 82 | 180 GB | $65.523 | $8.192 | Public USD GPU Clusters list | 2026-11-11 |
| gpu-b200-sxm / 8gpu-160vcpu-1792gb | Nebius | NVIDIA B200 | 81 | 180 GB | $68.002 | $8.501 | us-central1 | 2026-11-11 |
| b200-8x | CoreWeave | NVIDIA B200 | 81 | 180 GB | $68.802 | $8.601 | North America | 2026-11-11 |
| BM.GPU.MI355X.8 | Oracle | AMD MI355X | 81 | 288 GB | $68.802 | $8.602 | OCI USD list | 2026-11-11 |
| b300-288gb-sxm-ib.8x | Crusoe | NVIDIA B300 | 81 | 288 GB | $71.684 | $8.961 | Crusoe public USD GPU-hour list | 2026-11-11 |
| BM.GPU.GB300.4 | Oracle | NVIDIA GB300 | 41 | 278 GB | $72.002 | $18.002 | OCI USD list | 2026-11-11 |
| BM.GPU.H100.8 | Oracle | NVIDIA H100 | 81 | 80 GB | $80.002 | $10.001 | OCI USD list | 2026-11-11 |
| BM.GPU.H200.8 | Oracle | NVIDIA H200 | 81 | 141 GB | $80.002 | $10.001 | OCI USD list | 2026-11-11 |
| a3-ultragpu-8g | Google Cloud | NVIDIA H200 | 81 | 141 GB | $84.80692 | $10.60091 | us-central1 | 2026-11-11 |
| gx3d-160x1792x8h100 | IBM Cloud | NVIDIA H100 | 81 | 80 GB | $85.852 | $10.73132 | us-east | 2026-11-11 |
| a3-highgpu-8g | Google Cloud | NVIDIA H100 SXM | 81 | 80 GB | $88.492 | $11.06131 | us-central1 | 2026-11-11 |
| a3-megagpu-8g | Google Cloud | NVIDIA H100 | 81 | 80 GB | $93.40072 | $11.67511 | us-central1 (Iowa) | 2026-11-11 |
| Standard_ND96isr_H100_v5 | Azure | NVIDIA H100 | 81 | 80 GB | $98.322 | $12.291 | eastus | 2026-11-11 |
| Standard_ND128isr_NDR_GB200_v6 | Azure | NVIDIA GB200 | 41 | 192 GiB | $108.162 | $27.041 | eastus | 2026-11-11 |
| BM.GPU.B200.8 | Oracle | NVIDIA B200 | 81 | 180 GB | $112.002 | $14.002 | OCI USD list | 2026-11-11 |
| p6-b200.48xlarge | AWS | NVIDIA B200 | 81 | 180 GB | $113.93283 | $14.24161 | us-east-1 | 2026-11-11 |
| BM.GPU.B300.8 | Oracle | NVIDIA B300 | 81 | 263 GB | $120.002 | $15.002 | OCI USD list | 2026-11-11 |
| p6-b300.48xlarge | AWS | NVIDIA B300 | 81 | 268 GiB | $142.4162 | $17.8021 | us-east-1 | 2026-11-11 |
| a3-highgpu-1g | Google Cloud | NVIDIA H100 | 11 | 80 GB | — | — | — | 2026-11-11 |
| a3-highgpu-2g | Google Cloud | NVIDIA H100 | 21 | 80 GB | — | — | — | 2026-11-11 |
| a3-highgpu-4g | Google Cloud | NVIDIA H100 | 41 | 80 GB | — | — | — | 2026-11-11 |
| a4-highgpu-8g | Google Cloud | NVIDIA B200 | 81 | 180 GB | — | — | us-central1 | 2026-11-11 |
| b200-180gb-sxm-ib-8x | Crusoe | NVIDIA B200 | 81 | 180 GB | — | — | — | 2026-11-11 |
| B300-SXM-8-288G | Scaleway | NVIDIA B300-SXM | 81 | 288 GB | — | — | — | 2026-11-11 |
| BM.GPU2.2 | Oracle | NVIDIA P100 | 21 | 16 GB | — | — | — | 2026-11-11 |
| BM.GPU3.8 | Oracle | NVIDIA V100 | 81 | 16 GB | — | — | — | 2026-11-11 |
| ecs.gn7-c12g1.3xlarge | Alibaba | — | 11 | 40 GB | — | — | — | 2026-11-11 |
| ecs.gn7-c13g1.13xlarge | Alibaba | — | 41 | 40 GB | — | — | — | 2026-11-11 |
| ecs.gn7-c13g1.26xlarge | Alibaba | — | 81 | 40 GB | — | — | — | 2026-11-11 |
| ecs.gn7e-c16g1.16xlarge | Alibaba | — | 41 | 80 GB | — | — | — | 2026-11-11 |
| ecs.gn7e-c16g1.32xlarge | Alibaba | — | 81 | 80 GB | — | — | — | 2026-11-11 |
| ecs.gn7e-c16g1.4xlarge | Alibaba | — | 11 | 80 GB | — | — | — | 2026-11-11 |
| ecs.gn7e-c16g1.8xlarge | Alibaba | — | 21 | 80 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-2x.8xlarge | Alibaba | NVIDIA A10 | 21 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-4x.16xlarge | Alibaba | NVIDIA A10 | 41 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-4x.8xlarge | Alibaba | NVIDIA A10 | 41 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-8x.16xlarge | Alibaba | NVIDIA A10 | 81 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-8x.32xlarge | Alibaba | NVIDIA A10 | 81 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-c16g1.4xlarge | Alibaba | NVIDIA A10 | 11 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-c32g1.16xlarge | Alibaba | NVIDIA A10 | 21 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-c32g1.32xlarge | Alibaba | NVIDIA A10 | 41 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-c32g1.8xlarge | Alibaba | NVIDIA A10 | 11 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-c48g1.12xlarge | Alibaba | NVIDIA A10 | 11 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-c56g1.14xlarge | Alibaba | NVIDIA A10 | 11 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn7i-c8g1.2xlarge | Alibaba | NVIDIA A10 | 11 | 24 GB | — | — | — | 2026-11-11 |
| ecs.gn8is-2x.8xlarge | Alibaba | NVIDIA L20 | 21 | 48 GB | — | — | — | 2026-11-11 |
| ecs.gn8is-4x.16xlarge | Alibaba | NVIDIA L20 | 41 | 48 GB | — | — | — | 2026-11-11 |
| ecs.gn8is-8x.32xlarge | Alibaba | NVIDIA L20 | 81 | 48 GB | — | — | — | 2026-11-11 |
| ecs.gn8is.2xlarge | Alibaba | NVIDIA L20 | 11 | 48 GB | — | — | — | 2026-11-11 |
| ecs.gn8is.4xlarge | Alibaba | NVIDIA L20 | 11 | 48 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-2x.12xlarge | Alibaba | — | 21 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-2x.8xlarge | Alibaba | — | 21 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-4x.24xlarge | Alibaba | — | 41 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-4x.8xlarge | Alibaba | — | 41 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-8x.16xlarge | Alibaba | — | 81 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-8x.48xlarge | Alibaba | — | 81 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-tee-8x.16xlarge | Alibaba | — | 81 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-tee-8x.48xlarge | Alibaba | — | 81 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-tee.4xlarge | Alibaba | — | 11 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v-tee.6xlarge | Alibaba | — | 11 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v.4xlarge | Alibaba | — | 11 | 96 GB | — | — | — | 2026-11-11 |
| ecs.gn8v.6xlarge | Alibaba | — | 11 | 96 GB | — | — | — | 2026-11-11 |
| g6.10xlarge.7 | Huawei | NVIDIA T4 | 11 | 16 GiB | — | — | — | 2026-11-11 |
| g6.18xlarge.7 | Huawei | NVIDIA T4 | 21 | 16 GiB | — | — | — | 2026-11-11 |
| g6.20xlarge.7 | Huawei | NVIDIA T4 | 21 | 16 GiB | — | — | — | 2026-11-11 |
| g6.4xlarge.4 | Huawei | NVIDIA T4 | 11 | 16 GiB | — | — | — | 2026-11-11 |
| g6.6xlarge.4 | Huawei | NVIDIA T4 | 11 | 16 GiB | — | — | — | 2026-11-11 |
| g6.9xlarge.7 | Huawei | NVIDIA T4 | 11 | 16 GiB | — | — | — | 2026-11-11 |
| g6.xlarge.4 | Huawei | NVIDIA T4 | 11 | 16 GiB | — | — | — | 2026-11-11 |
| gb200-186gb-nvl-4x | Crusoe | NVIDIA GB200 | 41 | 186 GB | — | — | — | 2026-11-11 |
| GEX131 | Hetzner | NVIDIA RTX PRO 6000 Blackwell Max-Q | 11 | 96 GB | — | — | European dedicated-server EUR list | 2026-11-11 |
| GEX430 | Hetzner | NVIDIA RTX PRO 6000 Blackwell Max-Q | 41 | 96 GB | — | — | European dedicated-server EUR list | 2026-11-11 |
| GEX45 | Hetzner | NVIDIA RTX PRO 4000 Blackwell SFF | 11 | 24 GB | — | — | European dedicated-server EUR list | 2026-11-11 |
| GEX63 | Hetzner | NVIDIA RTX PRO 6000 Blackwell Max-Q | 11 | 96 GB | — | — | European dedicated-server EUR list | 2026-11-11 |
| GN10Xp.10XLARGE160 | Tencent | NVIDIA V100 | 41 | 32 GB | — | — | — | 2026-11-11 |
| GN10Xp.20XLARGE320 | Tencent | NVIDIA V100 | 81 | 32 GB | — | — | — | 2026-11-11 |
| GN10Xp.2XLARGE40 | Tencent | NVIDIA V100 | 11 | 32 GB | — | — | — | 2026-11-11 |
| GN10Xp.5XLARGE80 | Tencent | NVIDIA V100 | 21 | 32 GB | — | — | — | 2026-11-11 |
| GN7.10XLARGE160 | Tencent | NVIDIA T4 | 21 | 16 GB | — | — | — | 2026-11-11 |
| GN7.20XLARGE320 | Tencent | NVIDIA T4 | 41 | 16 GB | — | — | — | 2026-11-11 |
| GN7.2XLARGE32 | Tencent | NVIDIA T4 | 11 | 16 GB | — | — | — | 2026-11-11 |
| GN7.5XLARGE80 | Tencent | NVIDIA T4 | 11 | 16 GB | — | — | — | 2026-11-11 |
| GN7.8XLARGE128 | Tencent | NVIDIA T4 | 11 | 16 GB | — | — | — | 2026-11-11 |
| GT4.20XLARGE474 | Tencent | NVIDIA A100 | 41 | 40 GB | — | — | — | 2026-11-11 |
| GT4.41XLARGE948 | Tencent | NVIDIA A100 | 81 | 40 GB | — | — | — | 2026-11-11 |
| GT4.4XLARGE96 | Tencent | NVIDIA A100 | 11 | 40 GB | — | — | — | 2026-11-11 |
| GT4.8XLARGE192 | Tencent | NVIDIA A100 | 21 | 40 GB | — | — | — | 2026-11-11 |
| gx3-16x80x1l4 | IBM Cloud | NVIDIA L4 | 11 | 24 GB | — | — | — | 2026-11-11 |
| gx3-24x120x1l40s | IBM Cloud | NVIDIA L40S | 11 | 48 GB | — | — | — | 2026-11-11 |
| gx3-32x160x2l4 | IBM Cloud | NVIDIA L4 | 21 | 24 GB | — | — | — | 2026-11-11 |
| gx3-48x240x-2l40s | IBM Cloud | NVIDIA L40S | 21 | 48 GB | — | — | — | 2026-11-11 |
| gx3-64x320x4l4 | IBM Cloud | NVIDIA L4 | 41 | 24 GB | — | — | — | 2026-11-11 |
| gx3d-160x1792x8h200 | IBM Cloud | NVIDIA H200 | 81 | 141 GB | — | — | — | 2026-11-11 |
| gx3d-24x120x1a100p | IBM Cloud | NVIDIA A100 PCIe | 11 | 80 GB | — | — | — | 2026-11-11 |
| gx3d-48x240x2a100p | IBM Cloud | NVIDIA A100 PCIe | 21 | 80 GB | — | — | — | 2026-11-11 |
| H100-1-80G | Scaleway | NVIDIA H100-PCIe | 11 | 80 GB | — | — | — | 2026-11-11 |
| H100-2-80G | Scaleway | NVIDIA H100-PCIe | 21 | 80 GB | — | — | — | 2026-11-11 |
| H100-SXM-2-80G | Scaleway | NVIDIA H100-SXM | 21 | 80 GB | — | — | — | 2026-11-11 |
| H100-SXM-4-80G | Scaleway | NVIDIA H100-SXM | 41 | 80 GB | — | — | — | 2026-11-11 |
| H100-SXM-8-80G | Scaleway | NVIDIA H100-SXM | 81 | 80 GB | — | — | — | 2026-11-11 |
| L4-1-24G | Scaleway | NVIDIA L4 | 11 | 24 GB | — | — | fr-par-1 | 2026-11-11 |
| L4-2-24G | Scaleway | NVIDIA L4 | 21 | 24 GB | — | — | fr-par-1 | 2026-11-11 |
| L4-4-24G | Scaleway | NVIDIA L4 | 41 | 24 GB | — | — | fr-par-1 | 2026-11-11 |
| L4-8-24G | Scaleway | NVIDIA L4 | 81 | 24 GB | — | — | fr-par-1 | 2026-11-11 |
| L40S-1-48G | Scaleway | NVIDIA L40S | 11 | 48 GB | — | — | — | 2026-11-11 |
| L40S-2-48G | Scaleway | NVIDIA L40S | 21 | 48 GB | — | — | — | 2026-11-11 |
| L40S-4-48G | Scaleway | NVIDIA L40S | 41 | 48 GB | — | — | — | 2026-11-11 |
| L40S-8-48G | Scaleway | NVIDIA L40S | 81 | 48 GB | — | — | — | 2026-11-11 |
| mi355x-288gb-roce.8x | Crusoe | AMD MI355X | 81 | 288 GB | — | — | — | 2026-11-11 |
| p2vs.16xlarge.8 | Huawei | NVIDIA V100 | 81 | 32 GiB | — | — | — | 2026-11-11 |
| p2vs.2xlarge.8 | Huawei | NVIDIA V100 | 11 | 32 GiB | — | — | — | 2026-11-11 |
| p2vs.4xlarge.8 | Huawei | NVIDIA V100 | 21 | 32 GiB | — | — | — | 2026-11-11 |
| p2vs.8xlarge.8 | Huawei | NVIDIA V100 | 41 | 32 GiB | — | — | — | 2026-11-11 |
| p5e.48xlarge | AWS | NVIDIA H200 | 81 | 141 GB | — | — | us-east-1 | 2026-11-11 |
| p6e-gb200.36xlarge | AWS | NVIDIA B200 | 41 | 185 GiB | — | — | — | 2026-11-11 |
| pi2.2xlarge.4 | Huawei | NVIDIA T4 | 11 | 16 GiB | — | — | — | 2026-11-11 |
| pi2.4xlarge.4 | Huawei | NVIDIA T4 | 21 | 16 GiB | — | — | — | 2026-11-11 |
| pi2.8xlarge.4 | Huawei | NVIDIA T4 | 41 | 16 GiB | — | — | — | 2026-11-11 |
| PNV4.14XLARGE232 | Tencent | NVIDIA A10 | 21 | 24 GB | — | — | — | 2026-11-11 |
| PNV4.28XLARGE466 | Tencent | NVIDIA A10 | 41 | 24 GB | — | — | — | 2026-11-11 |
| PNV4.56XLARGE932 | Tencent | NVIDIA A10 | 81 | 24 GB | — | — | — | 2026-11-11 |
| PNV4.7XLARGE116 | Tencent | NVIDIA A10 | 11 | 24 GB | — | — | — | 2026-11-11 |
| Standard_ND96isr_H200_v5 | Azure | NVIDIA H200 | 81 | 141 GB | — | — | — | 2026-11-11 |
| Standard_ND96isr_MI300X_v5 | Azure | AMD MI300X | 81 | 192 GB | — | — | — | 2026-11-11 |
| vcpu-12_memory-72g_nvidia-rtx-3090-3 | Genesis Cloud | NVIDIA RTX 3090 | 31 | — | — | — | — | 2026-11-11 |
| vcpu-16_memory-96g_nvidia-rtx-3090-2 | Genesis Cloud | NVIDIA RTX 3090 | 21 | — | — | — | — | 2026-11-11 |
| vcpu-16_memory-96g_nvidia-rtx-3090-4 | Genesis Cloud | NVIDIA RTX 3090 | 41 | — | — | — | — | 2026-11-11 |
| vcpu-192_memory-1920g_nvidia-h100-sxm5-8 | Genesis Cloud | NVIDIA H100 SXM | 81 | 80 GB | — | — | — | 2026-11-11 |
| vcpu-20_memory-120g_nvidia-rtx-3090-5 | Genesis Cloud | NVIDIA RTX 3090 | 51 | — | — | — | — | 2026-11-11 |
| vcpu-24_memory-144g_nvidia-rtx-3090-3 | Genesis Cloud | NVIDIA RTX 3090 | 31 | — | — | — | — | 2026-11-11 |
| vcpu-24_memory-144g_nvidia-rtx-3090-6 | Genesis Cloud | NVIDIA RTX 3090 | 61 | — | — | — | — | 2026-11-11 |
| vcpu-24_memory-240g_nvidia-h100-sxm5-1 | Genesis Cloud | NVIDIA H100 SXM | 11 | 80 GB | — | — | — | 2026-11-11 |
| vcpu-28_memory-168g_nvidia-rtx-3090-7 | Genesis Cloud | NVIDIA RTX 3090 | 71 | — | — | — | — | 2026-11-11 |
| vcpu-32_memory-192g_nvidia-rtx-3090-4 | Genesis Cloud | NVIDIA RTX 3090 | 41 | — | — | — | — | 2026-11-11 |
| vcpu-32_memory-192g_nvidia-rtx-3090-8 | Genesis Cloud | NVIDIA RTX 3090 | 81 | — | — | — | — | 2026-11-11 |
| vcpu-4_memory-24g_nvidia-rtx-3090-1 | Genesis Cloud | NVIDIA RTX 3090 | 11 | — | — | — | — | 2026-11-11 |
| vcpu-8_memory-48g_nvidia-rtx-3090-1 | Genesis Cloud | NVIDIA RTX 3090 | 11 | — | — | — | — | 2026-11-11 |
| vcpu-8_memory-48g_nvidia-rtx-3090-2 | Genesis Cloud | NVIDIA RTX 3090 | 21 | — | — | — | — | 2026-11-11 |
USD per whole instance-hour, in the stated rate scope. Memory per GPU retains the provider’s GB or GiB label. GPU / hour is instance price divided by GPU count. Empty prices have an explanation on the instance page. Read price and memory definitions. Select up to four rows to compare.
Keep the purchase terms attached
MethodologyThe same GPU model ships in different memory sizes and configurations. Open an instance for its host RAM, network limits, price terms and sources, then compare alternatives with the same GPU count or memory per device.
The table uses current public on-demand rates. A provider’s GPU-hour quote is multiplied by the instance’s GPU count. Each number links to its source; the checked date records when that source was read.
JSON with field-level sourcesAgent indexAbout the publisherNeocloud companiesHPC providers
Questions before you rent
Is the price for one GPU or the whole instance?
The Instance / hour column is the whole-instance USD rate. When a provider quotes a GPU-hour, we multiply it by the GPU count. Each detail page shows the calculation and purchase terms. Spot, reservation, Flex-start and inference-only prices are separate offers.
Does more GPU memory mean the model fits on one GPU?
Compare memory per GPU first. The total across an instance is split across its devices. Running a model across devices requires suitable software and communication between GPUs. Host RAM is a separate field on each detail page.
What do EFA, InfiniBand, NVLink and NVSwitch describe?
NVLink and NVSwitch connect GPUs within a system. EFA, InfiniBand and RDMA describe networking used to communicate between nodes. The interconnect and network fields preserve those distinctions instead of combining their bandwidth figures.
Why are some hourly prices empty?
An empty USD cell means the provider publishes no public on-demand USD rate for that exact configuration. Google A4 sells through Spot, Flex-start and reservations; many Asian-cloud shapes price only in regional calculators. Hetzner and Scaleway quote in euros, shown on their instance pages. Lambda rows carry its public list, effective 16 November 2026.
Does this cover every cloud and region?
It covers 351 full-GPU configurations across 22 providers: hyperscaler SKUs, neocloud presets and dedicated servers, with published regions on each instance page. Live stock and account quota change by the hour, so confirm capacity in the provider console before you commit. The coverage page lists what each provider contributes.
See Google A4’s purchase model, Lambda’s 16 November list and Hetzner’s euro pricing for the cases above.