# Cloud GPU List

> 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.

Read the spec. Rent the right GPU. 351 instances, each field linked to its public source, last verified 2026-11-11.
When citing, link the canonical HTML page.

## Data
- [Cloud GPU instances](https://cloudgpulist.com/): the full comparison table
- [Data (JSON)](https://cloudgpulist.com/data.json): every record with sources
- [Data (CSV)](https://cloudgpulist.com/data.csv): one row per instance
- [Methodology](https://cloudgpulist.com/methodology/): what each field means and how it is verified

## Instances
- [1x A10](https://cloudgpulist.com/lambda-1x-a10/): Lambda 1x A10 allocates 1 NVIDIA A10 accelerator.
- [1x A100 PCIe](https://cloudgpulist.com/lambda-1x-a100-pcie/): Lambda 1x A100 PCIe supplies 1 NVIDIA A100 PCIe GPU.
- [1x A100 SXM](https://cloudgpulist.com/lambda-1x-a100-sxm/): Lambda 1x A100 SXM allocates 1 NVIDIA A100 SXM accelerator.
- [1x A6000](https://cloudgpulist.com/lambda-1x-a6000/): Lambda 1x A6000 supplies 1 NVIDIA A6000 GPU.
- [1x B200 SXM6](https://cloudgpulist.com/lambda-1x-b200-sxm6/): Lambda 1x B200 SXM6 supplies 1 NVIDIA B200 SXM6 GPU.
- [1x GH200](https://cloudgpulist.com/lambda-1x-gh200/): Lambda 1x GH200 allocates 1 NVIDIA GH200 accelerator.
- [1x H100 PCIe](https://cloudgpulist.com/lambda-1x-h100-pcie/): Lambda 1x H100 PCIe allocates 1 NVIDIA H100 PCIe accelerator.
- [1x H100 SXM](https://cloudgpulist.com/lambda-1x-h100-sxm/): Lambda 1x H100 SXM supplies 1 NVIDIA H100 SXM GPU.
- [1x Quadro RTX 6000](https://cloudgpulist.com/lambda-1x-quadro-rtx-6000/): Lambda 1x Quadro RTX 6000 allocates 1 NVIDIA Quadro RTX 6000 accelerator.
- [2x A100 PCIe](https://cloudgpulist.com/lambda-2x-a100-pcie/): Lambda 2x A100 PCIe allocates 2 NVIDIA A100 PCIe accelerators.
- [2x A6000](https://cloudgpulist.com/lambda-2x-a6000/): Lambda 2x A6000 allocates 2 NVIDIA A6000 accelerators.
- [2x B200 SXM6](https://cloudgpulist.com/lambda-2x-b200-sxm6/): Lambda 2x B200 SXM6 allocates 2 NVIDIA B200 SXM6 accelerators.
- [2x H100 SXM](https://cloudgpulist.com/lambda-2x-h100-sxm/): Lambda 2x H100 SXM allocates 2 NVIDIA H100 SXM accelerators.
- [4x A100 PCIe](https://cloudgpulist.com/lambda-4x-a100-pcie/): Lambda 4x A100 PCIe allocates 4 NVIDIA A100 PCIe accelerators.
- [4x A6000](https://cloudgpulist.com/lambda-4x-a6000/): Lambda 4x A6000 allocates 4 NVIDIA A6000 accelerators.
- [4x B200 SXM6](https://cloudgpulist.com/lambda-4x-b200-sxm6/): Lambda 4x B200 SXM6 allocates 4 NVIDIA B200 SXM6 accelerators.
- [4x H100 SXM](https://cloudgpulist.com/lambda-4x-h100-sxm/): Lambda 4x H100 SXM allocates 4 NVIDIA H100 SXM accelerators.
- [8x HGX A100 40GB](https://cloudgpulist.com/lambda-8x-hgx-a100-40gb/): Lambda 8x HGX A100 40GB supplies eight NVIDIA HGX A100 GPUs with 40 GB per device.
- [8x HGX A100 80GB](https://cloudgpulist.com/lambda-8x-hgx-a100-80gb/): Lambda 8x HGX A100 80GB supplies eight NVIDIA HGX A100 GPUs with 80 GB per device.
- [8x HGX B200](https://cloudgpulist.com/lambda-8x-hgx-b200/): Lambda 8x HGX B200 supplies 8 NVIDIA HGX B200 GPUs.
- [8x HGX H100](https://cloudgpulist.com/lambda-8x-hgx-h100/): Lambda 8x HGX H100 supplies 8 NVIDIA HGX H100 GPUs.
- [8x Tesla V100](https://cloudgpulist.com/lambda-8x-tesla-v100/): Lambda 8x Tesla V100 allocates 8 NVIDIA Tesla V100 accelerators.
- [a10-180](https://cloudgpulist.com/ovhcloud-a10-180/): OVHcloud a10-180 allocates 4 NVIDIA A10 accelerators.
- [a10-45](https://cloudgpulist.com/ovhcloud-a10-45/): OVHcloud a10-45 allocates 1 NVIDIA A10 accelerator.
- [a10-90](https://cloudgpulist.com/ovhcloud-a10-90/): OVHcloud a10-90 allocates 2 NVIDIA A10 accelerators.
- [A100](https://cloudgpulist.com/paperspace-a100/): Paperspace A100 allocates 1 NVIDIA A100 accelerator.
- [A100 PCIe / 1 GPU Pod](https://cloudgpulist.com/runpod-a100-pcie-1-gpu-pod/): RunPod A100 PCIe / 1 GPU Pod allocates 1 NVIDIA A100 PCIe accelerator.
- [A100 SXM / 1 GPU Pod](https://cloudgpulist.com/runpod-a100-sxm-1-gpu-pod/): RunPod A100 SXM / 1 GPU Pod allocates 1 NVIDIA A100 SXM accelerator.
- [a100-180](https://cloudgpulist.com/ovhcloud-a100-180/): OVHcloud a100-180 allocates 1 NVIDIA A100 accelerator.
- [a100-360](https://cloudgpulist.com/ovhcloud-a100-360/): OVHcloud a100-360 allocates 2 NVIDIA A100 accelerators.
- [a100-720](https://cloudgpulist.com/ovhcloud-a100-720/): OVHcloud a100-720 allocates 4 NVIDIA A100 accelerators.
- [A100-80G](https://cloudgpulist.com/paperspace-a100-80g/): Paperspace A100-80G allocates 1 NVIDIA A100 accelerator.
- [a100-80gb-sxm-ib.8x](https://cloudgpulist.com/crusoe-a100-80gb-sxm-ib-8x/): Crusoe a100-80gb-sxm-ib.8x allocates 8 NVIDIA A100 accelerators.
- [a100-80gb.1x](https://cloudgpulist.com/crusoe-a100-80gb-1x/): Crusoe a100-80gb.1x allocates 1 NVIDIA A100 accelerator.
- [a100-80gb.2x](https://cloudgpulist.com/crusoe-a100-80gb-2x/): Crusoe a100-80gb.2x allocates 2 NVIDIA A100 accelerators.
- [a100-80gb.4x](https://cloudgpulist.com/crusoe-a100-80gb-4x/): Crusoe a100-80gb.4x allocates 4 NVIDIA A100 accelerators.
- [a100-80gb.8x](https://cloudgpulist.com/crusoe-a100-80gb-8x/): Crusoe a100-80gb.8x allocates 8 NVIDIA A100 accelerators.
- [a2-highgpu-1g](https://cloudgpulist.com/google-cloud-a2-highgpu-1g/): Google Cloud a2-highgpu-1g supplies 1 NVIDIA A100 GPU.
- [a2-highgpu-8g](https://cloudgpulist.com/google-cloud-a2-highgpu-8g/): Google Cloud a2-highgpu-8g supplies 8 NVIDIA A100 GPUs.
- [a2-megagpu-16g](https://cloudgpulist.com/google-cloud-a2-megagpu-16g/): Google Cloud a2-megagpu-16g supplies 16 NVIDIA A100 GPUs.
- [a2-ultragpu-1g](https://cloudgpulist.com/google-cloud-a2-ultragpu-1g/): Google Cloud a2-ultragpu-1g supplies 1 NVIDIA A100 GPU.
- [a2-ultragpu-8g](https://cloudgpulist.com/google-cloud-a2-ultragpu-8g/): Google Cloud a2-ultragpu-8g supplies 8 NVIDIA A100 GPUs.
- [a3-highgpu-1g](https://cloudgpulist.com/google-cloud-a3-highgpu-1g/): Google Cloud a3-highgpu-1g allocates 1 NVIDIA H100 accelerator.
- [a3-highgpu-2g](https://cloudgpulist.com/google-cloud-a3-highgpu-2g/): Google Cloud a3-highgpu-2g allocates 2 NVIDIA H100 accelerators.
- [a3-highgpu-4g](https://cloudgpulist.com/google-cloud-a3-highgpu-4g/): Google Cloud a3-highgpu-4g allocates 4 NVIDIA H100 accelerators.
- [a3-highgpu-8g](https://cloudgpulist.com/google-cloud-a3-highgpu-8g/): Google Cloud a3-highgpu-8g supplies 8 NVIDIA H100 SXM GPUs.
- [a3-megagpu-8g](https://cloudgpulist.com/google-cloud-a3-megagpu-8g/): Google Cloud a3-megagpu-8g allocates 8 NVIDIA H100 accelerators.
- [a3-ultragpu-8g](https://cloudgpulist.com/google-cloud-a3-ultragpu-8g/): Google Cloud a3-ultragpu-8g supplies 8 NVIDIA H200 GPUs.
- [a4-highgpu-8g](https://cloudgpulist.com/google-cloud-a4-highgpu-8g/): Google Cloud a4-highgpu-8g supplies 8 NVIDIA B200 GPUs.
- [A40 / 1 GPU Pod](https://cloudgpulist.com/runpod-a40-1-gpu-pod/): RunPod A40 / 1 GPU Pod allocates 1 NVIDIA A40 accelerator.
- [A4000](https://cloudgpulist.com/paperspace-a4000/): Paperspace A4000 allocates 1 NVIDIA A4000 accelerator.
- [A5000](https://cloudgpulist.com/paperspace-a5000/): Paperspace A5000 allocates 1 NVIDIA A5000 accelerator.
- [A6000](https://cloudgpulist.com/paperspace-a6000/): Paperspace A6000 allocates 1 NVIDIA A6000 accelerator.
- [B200 / 1 GPU Pod](https://cloudgpulist.com/runpod-b200-1-gpu-pod/): RunPod B200 / 1 GPU Pod allocates 1 NVIDIA B200 accelerator.
- [b200-180gb-sxm-ib-8x](https://cloudgpulist.com/crusoe-b200-180gb-sxm-ib-8x/): Crusoe b200-180gb-sxm-ib-8x allocates 8 NVIDIA B200 accelerators.
- [b200-8x](https://cloudgpulist.com/coreweave-b200-8x/): CoreWeave b200-8x supplies 8 NVIDIA B200 GPUs.
- [B300 / 1 GPU Pod](https://cloudgpulist.com/runpod-b300-1-gpu-pod/): RunPod B300 / 1 GPU Pod allocates 1 NVIDIA B300 accelerator.
- [b300-288gb-sxm-ib.8x](https://cloudgpulist.com/crusoe-b300-288gb-sxm-ib-8x/): Crusoe b300-288gb-sxm-ib.8x allocates 8 NVIDIA B300 accelerators.
- [B300-SXM-8-288G](https://cloudgpulist.com/scaleway-b300-sxm-8-288g/): Scaleway B300-SXM-8-288G allocates 8 NVIDIA B300-SXM accelerators.
- [BM.GPU.A10.4](https://cloudgpulist.com/oracle-bm-gpu-a10-4/): Oracle BM.GPU.A10.4 supplies 4 NVIDIA A10 GPUs.
- [BM.GPU.A100-v2.8](https://cloudgpulist.com/oracle-bm-gpu-a100-v2-8/): Oracle BM.GPU.A100-v2.8 supplies 8 NVIDIA A100 GPUs.
- [BM.GPU.B200.8](https://cloudgpulist.com/oracle-bm-gpu-b200-8/): Oracle BM.GPU.B200.8 supplies 8 NVIDIA B200 GPUs.
- [BM.GPU.B300.8](https://cloudgpulist.com/oracle-bm-gpu-b300-8/): Oracle BM.GPU.B300.8 allocates 8 NVIDIA B300 accelerators.
- [BM.GPU.GB200.4](https://cloudgpulist.com/oracle-bm-gpu-gb200-4/): Oracle BM.GPU.GB200.4 allocates 4 NVIDIA GB200 accelerators.
- [BM.GPU.GB300.4](https://cloudgpulist.com/oracle-bm-gpu-gb300-4/): Oracle BM.GPU.GB300.4 allocates 4 NVIDIA GB300 accelerators.
- [BM.GPU.H100.8](https://cloudgpulist.com/oracle-bm-gpu-h100-8/): Oracle BM.GPU.H100.8 supplies 8 NVIDIA H100 GPUs.
- [BM.GPU.H200.8](https://cloudgpulist.com/oracle-bm-gpu-h200-8/): Oracle BM.GPU.H200.8 supplies 8 NVIDIA H200 GPUs.
- [BM.GPU.L40S.4](https://cloudgpulist.com/oracle-bm-gpu-l40s-4/): Oracle BM.GPU.L40S.4 allocates 4 NVIDIA L40S accelerators.
- [BM.GPU.MI300X.8](https://cloudgpulist.com/oracle-bm-gpu-mi300x-8/): Oracle BM.GPU.MI300X.8 supplies 8 AMD MI300X GPUs.
- [BM.GPU.MI355X.8](https://cloudgpulist.com/oracle-bm-gpu-mi355x-8/): Oracle BM.GPU.MI355X.8 allocates 8 AMD MI355X accelerators.
- [BM.GPU.RTXPRO.8](https://cloudgpulist.com/oracle-bm-gpu-rtxpro-8/): Oracle BM.GPU.RTXPRO.8 allocates 8 NVIDIA RTX PRO 6000 accelerators.
- [BM.GPU2.2](https://cloudgpulist.com/oracle-bm-gpu2-2/): Oracle BM.GPU2.2 allocates 2 NVIDIA P100 accelerators.
- [BM.GPU3.8](https://cloudgpulist.com/oracle-bm-gpu3-8/): Oracle BM.GPU3.8 allocates 8 NVIDIA V100 accelerators.
- [BM.GPU4.8](https://cloudgpulist.com/oracle-bm-gpu4-8/): Oracle BM.GPU4.8 supplies 8 NVIDIA A100 GPUs.
- [ecs.gn7-c12g1.3xlarge](https://cloudgpulist.com/alibaba-ecs-gn7-c12g1-3xlarge/): Alibaba ecs.gn7-c12g1.3xlarge allocates 1 GPUs; the cited specification does not identify the model.
- [ecs.gn7-c13g1.13xlarge](https://cloudgpulist.com/alibaba-ecs-gn7-c13g1-13xlarge/): Alibaba ecs.gn7-c13g1.13xlarge allocates 4 GPUs; the cited specification does not identify the model.
- [ecs.gn7-c13g1.26xlarge](https://cloudgpulist.com/alibaba-ecs-gn7-c13g1-26xlarge/): Alibaba ecs.gn7-c13g1.26xlarge allocates 8 GPUs; the cited specification does not identify the model.
- [ecs.gn7e-c16g1.16xlarge](https://cloudgpulist.com/alibaba-ecs-gn7e-c16g1-16xlarge/): Alibaba ecs.gn7e-c16g1.16xlarge allocates 4 GPUs; the cited specification does not identify the model.
- [ecs.gn7e-c16g1.32xlarge](https://cloudgpulist.com/alibaba-ecs-gn7e-c16g1-32xlarge/): Alibaba ecs.gn7e-c16g1.32xlarge allocates 8 GPUs; the cited specification does not identify the model.
- [ecs.gn7e-c16g1.4xlarge](https://cloudgpulist.com/alibaba-ecs-gn7e-c16g1-4xlarge/): Alibaba ecs.gn7e-c16g1.4xlarge allocates 1 GPUs; the cited specification does not identify the model.
- [ecs.gn7e-c16g1.8xlarge](https://cloudgpulist.com/alibaba-ecs-gn7e-c16g1-8xlarge/): Alibaba ecs.gn7e-c16g1.8xlarge allocates 2 GPUs; the cited specification does not identify the model.
- [ecs.gn7i-2x.8xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-2x-8xlarge/): Alibaba ecs.gn7i-2x.8xlarge allocates 2 NVIDIA A10 accelerators.
- [ecs.gn7i-4x.16xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-4x-16xlarge/): Alibaba ecs.gn7i-4x.16xlarge allocates 4 NVIDIA A10 accelerators.
- [ecs.gn7i-4x.8xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-4x-8xlarge/): Alibaba ecs.gn7i-4x.8xlarge allocates 4 NVIDIA A10 accelerators.
- [ecs.gn7i-8x.16xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-8x-16xlarge/): Alibaba ecs.gn7i-8x.16xlarge allocates 8 NVIDIA A10 accelerators.
- [ecs.gn7i-8x.32xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-8x-32xlarge/): Alibaba ecs.gn7i-8x.32xlarge allocates 8 NVIDIA A10 accelerators.
- [ecs.gn7i-c16g1.4xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-c16g1-4xlarge/): Alibaba ecs.gn7i-c16g1.4xlarge allocates 1 NVIDIA A10 accelerator.
- [ecs.gn7i-c32g1.16xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-c32g1-16xlarge/): Alibaba ecs.gn7i-c32g1.16xlarge allocates 2 NVIDIA A10 accelerators.
- [ecs.gn7i-c32g1.32xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-c32g1-32xlarge/): Alibaba ecs.gn7i-c32g1.32xlarge allocates 4 NVIDIA A10 accelerators.
- [ecs.gn7i-c32g1.8xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-c32g1-8xlarge/): Alibaba ecs.gn7i-c32g1.8xlarge allocates 1 NVIDIA A10 accelerator.
- [ecs.gn7i-c48g1.12xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-c48g1-12xlarge/): Alibaba ecs.gn7i-c48g1.12xlarge allocates 1 NVIDIA A10 accelerator.
- [ecs.gn7i-c56g1.14xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-c56g1-14xlarge/): Alibaba ecs.gn7i-c56g1.14xlarge allocates 1 NVIDIA A10 accelerator.
- [ecs.gn7i-c8g1.2xlarge](https://cloudgpulist.com/alibaba-ecs-gn7i-c8g1-2xlarge/): Alibaba ecs.gn7i-c8g1.2xlarge allocates 1 NVIDIA A10 accelerator.
- [ecs.gn8is-2x.8xlarge](https://cloudgpulist.com/alibaba-ecs-gn8is-2x-8xlarge/): Alibaba ecs.gn8is-2x.8xlarge allocates 2 NVIDIA L20 accelerators.
- [ecs.gn8is-4x.16xlarge](https://cloudgpulist.com/alibaba-ecs-gn8is-4x-16xlarge/): Alibaba ecs.gn8is-4x.16xlarge allocates 4 NVIDIA L20 accelerators.
- [ecs.gn8is-8x.32xlarge](https://cloudgpulist.com/alibaba-ecs-gn8is-8x-32xlarge/): Alibaba ecs.gn8is-8x.32xlarge allocates 8 NVIDIA L20 accelerators.
- [ecs.gn8is.2xlarge](https://cloudgpulist.com/alibaba-ecs-gn8is-2xlarge/): Alibaba ecs.gn8is.2xlarge allocates 1 NVIDIA L20 accelerator.
- [ecs.gn8is.4xlarge](https://cloudgpulist.com/alibaba-ecs-gn8is-4xlarge/): Alibaba ecs.gn8is.4xlarge allocates 1 NVIDIA L20 accelerator.
- [ecs.gn8v-2x.12xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-2x-12xlarge/): Alibaba ecs.gn8v-2x.12xlarge allocates 2 GPU accelerators.
- [ecs.gn8v-2x.8xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-2x-8xlarge/): Alibaba ecs.gn8v-2x.8xlarge allocates 2 GPU accelerators.
- [ecs.gn8v-4x.24xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-4x-24xlarge/): Alibaba ecs.gn8v-4x.24xlarge allocates 4 GPU accelerators.
- [ecs.gn8v-4x.8xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-4x-8xlarge/): Alibaba ecs.gn8v-4x.8xlarge allocates 4 GPU accelerators.
- [ecs.gn8v-8x.16xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-8x-16xlarge/): Alibaba ecs.gn8v-8x.16xlarge allocates 8 GPU accelerators.
- [ecs.gn8v-8x.48xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-8x-48xlarge/): Alibaba ecs.gn8v-8x.48xlarge allocates 8 GPU accelerators.
- [ecs.gn8v-tee-8x.16xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-tee-8x-16xlarge/): Alibaba ecs.gn8v-tee-8x.16xlarge allocates 8 GPU accelerators.
- [ecs.gn8v-tee-8x.48xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-tee-8x-48xlarge/): Alibaba ecs.gn8v-tee-8x.48xlarge allocates 8 GPU accelerators.
- [ecs.gn8v-tee.4xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-tee-4xlarge/): Alibaba ecs.gn8v-tee.4xlarge allocates 1 GPU accelerator.
- [ecs.gn8v-tee.6xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-tee-6xlarge/): Alibaba ecs.gn8v-tee.6xlarge allocates 1 GPU accelerator.
- [ecs.gn8v.4xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-4xlarge/): Alibaba ecs.gn8v.4xlarge allocates 1 GPU accelerator.
- [ecs.gn8v.6xlarge](https://cloudgpulist.com/alibaba-ecs-gn8v-6xlarge/): Alibaba ecs.gn8v.6xlarge allocates 1 GPU accelerator.
- [g2-standard-12](https://cloudgpulist.com/google-cloud-g2-standard-12/): Google Cloud g2-standard-12 allocates 1 NVIDIA L4 accelerator.
- [g2-standard-16](https://cloudgpulist.com/google-cloud-g2-standard-16/): Google Cloud g2-standard-16 allocates 1 NVIDIA L4 accelerator.
- [g2-standard-24](https://cloudgpulist.com/google-cloud-g2-standard-24/): Google Cloud g2-standard-24 supplies 2 NVIDIA L4 GPUs.
- [g2-standard-32](https://cloudgpulist.com/google-cloud-g2-standard-32/): Google Cloud g2-standard-32 allocates 1 NVIDIA L4 accelerator.
- [g2-standard-4](https://cloudgpulist.com/google-cloud-g2-standard-4/): Google Cloud g2-standard-4 supplies 1 NVIDIA L4 GPU.
- [g2-standard-48](https://cloudgpulist.com/google-cloud-g2-standard-48/): Google Cloud g2-standard-48 supplies 4 NVIDIA L4 GPUs.
- [g2-standard-8](https://cloudgpulist.com/google-cloud-g2-standard-8/): Google Cloud g2-standard-8 allocates 1 NVIDIA L4 accelerator.
- [g2-standard-96](https://cloudgpulist.com/google-cloud-g2-standard-96/): Google Cloud g2-standard-96 supplies 8 NVIDIA L4 GPUs.
- [g4dn.xlarge](https://cloudgpulist.com/aws-g4dn-xlarge/): AWS g4dn.xlarge supplies 1 NVIDIA T4 GPU.
- [g5.12xlarge](https://cloudgpulist.com/aws-g5-12xlarge/): AWS g5.12xlarge supplies 4 NVIDIA A10G GPUs.
- [g5.48xlarge](https://cloudgpulist.com/aws-g5-48xlarge/): AWS g5.48xlarge supplies 8 NVIDIA A10G GPUs.
- [g5.xlarge](https://cloudgpulist.com/aws-g5-xlarge/): AWS g5.xlarge supplies 1 NVIDIA A10G GPU.
- [g6.10xlarge.7](https://cloudgpulist.com/huawei-g6-10xlarge-7/): Huawei g6.10xlarge.7 allocates 1 NVIDIA T4 accelerator.
- [g6.12xlarge](https://cloudgpulist.com/aws-g6-12xlarge/): AWS g6.12xlarge supplies 4 NVIDIA L4 GPUs.
- [g6.16xlarge](https://cloudgpulist.com/aws-g6-16xlarge/): AWS g6.16xlarge allocates 1 NVIDIA L4 accelerator.
- [g6.18xlarge.7](https://cloudgpulist.com/huawei-g6-18xlarge-7/): Huawei g6.18xlarge.7 allocates 2 NVIDIA T4 accelerators.
- [g6.20xlarge.7](https://cloudgpulist.com/huawei-g6-20xlarge-7/): Huawei g6.20xlarge.7 allocates 2 NVIDIA T4 accelerators.
- [g6.24xlarge](https://cloudgpulist.com/aws-g6-24xlarge/): AWS g6.24xlarge allocates 4 NVIDIA L4 accelerators.
- [g6.2xlarge](https://cloudgpulist.com/aws-g6-2xlarge/): AWS g6.2xlarge allocates 1 NVIDIA L4 accelerator.
- [g6.48xlarge](https://cloudgpulist.com/aws-g6-48xlarge/): AWS g6.48xlarge allocates 8 NVIDIA L4 accelerators.
- [g6.4xlarge](https://cloudgpulist.com/aws-g6-4xlarge/): AWS g6.4xlarge allocates 1 NVIDIA L4 accelerator.
- [g6.4xlarge.4](https://cloudgpulist.com/huawei-g6-4xlarge-4/): Huawei g6.4xlarge.4 allocates 1 NVIDIA T4 accelerator.
- [g6.6xlarge.4](https://cloudgpulist.com/huawei-g6-6xlarge-4/): Huawei g6.6xlarge.4 allocates 1 NVIDIA T4 accelerator.
- [g6.8xlarge](https://cloudgpulist.com/aws-g6-8xlarge/): AWS g6.8xlarge allocates 1 NVIDIA L4 accelerator.
- [g6.9xlarge.7](https://cloudgpulist.com/huawei-g6-9xlarge-7/): Huawei g6.9xlarge.7 allocates 1 NVIDIA T4 accelerator.
- [g6.xlarge](https://cloudgpulist.com/aws-g6-xlarge/): AWS g6.xlarge supplies 1 NVIDIA L4 GPU.
- [g6.xlarge.4](https://cloudgpulist.com/huawei-g6-xlarge-4/): Huawei g6.xlarge.4 allocates 1 NVIDIA T4 accelerator.
- [g6e.12xlarge](https://cloudgpulist.com/aws-g6e-12xlarge/): AWS g6e.12xlarge supplies 4 NVIDIA L40S GPUs.
- [g6e.16xlarge](https://cloudgpulist.com/aws-g6e-16xlarge/): AWS g6e.16xlarge allocates 1 NVIDIA L40S accelerator.
- [g6e.24xlarge](https://cloudgpulist.com/aws-g6e-24xlarge/): AWS g6e.24xlarge allocates 4 NVIDIA L40S accelerators.
- [g6e.2xlarge](https://cloudgpulist.com/aws-g6e-2xlarge/): AWS g6e.2xlarge allocates 1 NVIDIA L40S accelerator.
- [g6e.48xlarge](https://cloudgpulist.com/aws-g6e-48xlarge/): AWS g6e.48xlarge allocates 8 NVIDIA L40S accelerators.
- [g6e.4xlarge](https://cloudgpulist.com/aws-g6e-4xlarge/): AWS g6e.4xlarge allocates 1 NVIDIA L40S accelerator.
- [g6e.8xlarge](https://cloudgpulist.com/aws-g6e-8xlarge/): AWS g6e.8xlarge allocates 1 NVIDIA L40S accelerator.
- [g6e.xlarge](https://cloudgpulist.com/aws-g6e-xlarge/): AWS g6e.xlarge supplies 1 NVIDIA L40S GPU.
- [g7e.2xlarge](https://cloudgpulist.com/aws-g7e-2xlarge/): AWS g7e.2xlarge supplies 1 NVIDIA RTX PRO 6000 GPU.
- [gb200-186gb-nvl-4x](https://cloudgpulist.com/crusoe-gb200-186gb-nvl-4x/): Crusoe gb200-186gb-nvl-4x allocates 4 NVIDIA GB200 accelerators.
- [gd-8xh100ib-i128](https://cloudgpulist.com/coreweave-gd-8xh100ib-i128/): CoreWeave gd-8xh100ib-i128 supplies 8 NVIDIA H100 GPUs.
- [gd-8xh200ib-i128](https://cloudgpulist.com/coreweave-gd-8xh200ib-i128/): CoreWeave gd-8xh200ib-i128 supplies 8 NVIDIA H200 GPUs.
- [gd-8xl40s-i128](https://cloudgpulist.com/coreweave-gd-8xl40s-i128/): CoreWeave gd-8xl40s-i128 supplies 8 NVIDIA L40S GPUs.
- [GEX131](https://cloudgpulist.com/hetzner-gex131/): Hetzner GEX131 allocates 1 NVIDIA RTX PRO 6000 Blackwell Max-Q accelerator.
- [GEX430](https://cloudgpulist.com/hetzner-gex430/): Hetzner GEX430 allocates 4 NVIDIA RTX PRO 6000 Blackwell Max-Q accelerators.
- [GEX45](https://cloudgpulist.com/hetzner-gex45/): Hetzner GEX45 allocates 1 NVIDIA RTX PRO 4000 Blackwell SFF accelerator.
- [GEX63](https://cloudgpulist.com/hetzner-gex63/): Hetzner GEX63 allocates 1 NVIDIA RTX PRO 6000 Blackwell Max-Q accelerator.
- [GN10Xp.10XLARGE160](https://cloudgpulist.com/tencent-gn10xp-10xlarge160/): Tencent GN10Xp.10XLARGE160 allocates 4 NVIDIA V100 accelerators.
- [GN10Xp.20XLARGE320](https://cloudgpulist.com/tencent-gn10xp-20xlarge320/): Tencent GN10Xp.20XLARGE320 allocates 8 NVIDIA V100 accelerators.
- [GN10Xp.2XLARGE40](https://cloudgpulist.com/tencent-gn10xp-2xlarge40/): Tencent GN10Xp.2XLARGE40 allocates 1 NVIDIA V100 accelerator.
- [GN10Xp.5XLARGE80](https://cloudgpulist.com/tencent-gn10xp-5xlarge80/): Tencent GN10Xp.5XLARGE80 allocates 2 NVIDIA V100 accelerators.
- [GN7.10XLARGE160](https://cloudgpulist.com/tencent-gn7-10xlarge160/): Tencent GN7.10XLARGE160 allocates 2 NVIDIA T4 accelerators.
- [GN7.20XLARGE320](https://cloudgpulist.com/tencent-gn7-20xlarge320/): Tencent GN7.20XLARGE320 allocates 4 NVIDIA T4 accelerators.
- [GN7.2XLARGE32](https://cloudgpulist.com/tencent-gn7-2xlarge32/): Tencent GN7.2XLARGE32 allocates 1 NVIDIA T4 accelerator.
- [GN7.5XLARGE80](https://cloudgpulist.com/tencent-gn7-5xlarge80/): Tencent GN7.5XLARGE80 allocates 1 NVIDIA T4 accelerator.
- [GN7.8XLARGE128](https://cloudgpulist.com/tencent-gn7-8xlarge128/): Tencent GN7.8XLARGE128 allocates 1 NVIDIA T4 accelerator.
- [gpu-4000adax1-20gb](https://cloudgpulist.com/digitalocean-gpu-4000adax1-20gb/): DigitalOcean gpu-4000adax1-20gb allocates 1 NVIDIA RTX 4000 Ada accelerator.
- [gpu-6000adax1-48gb](https://cloudgpulist.com/digitalocean-gpu-6000adax1-48gb/): DigitalOcean gpu-6000adax1-48gb allocates 1 NVIDIA RTX 6000 Ada accelerator.
- [gpu-b200-sxm / 1gpu-20vcpu-224gb](https://cloudgpulist.com/nebius-gpu-b200-sxm-1gpu-20vcpu-224gb/): Nebius gpu-b200-sxm / 1gpu-20vcpu-224gb supplies 1 NVIDIA B200 GPU.
- [gpu-b200-sxm / 8gpu-160vcpu-1792gb](https://cloudgpulist.com/nebius-gpu-b200-sxm-8gpu-160vcpu-1792gb/): Nebius gpu-b200-sxm / 8gpu-160vcpu-1792gb supplies 8 NVIDIA B200 GPUs.
- [gpu-h100-sxm / 1gpu-16vcpu-200gb](https://cloudgpulist.com/nebius-gpu-h100-sxm-1gpu-16vcpu-200gb/): Nebius gpu-h100-sxm / 1gpu-16vcpu-200gb supplies 1 NVIDIA H100 GPU.
- [gpu-h100-sxm / 8gpu-128vcpu-1600gb](https://cloudgpulist.com/nebius-gpu-h100-sxm-8gpu-128vcpu-1600gb/): Nebius gpu-h100-sxm / 8gpu-128vcpu-1600gb supplies 8 NVIDIA H100 GPUs.
- [gpu-h100x1-80gb](https://cloudgpulist.com/digitalocean-gpu-h100x1-80gb/): DigitalOcean gpu-h100x1-80gb allocates 1 NVIDIA H100 accelerator.
- [gpu-h100x8-640gb](https://cloudgpulist.com/digitalocean-gpu-h100x8-640gb/): DigitalOcean gpu-h100x8-640gb allocates 8 NVIDIA H100 accelerators.
- [gpu-h200-sxm / 1gpu-16vcpu-200gb](https://cloudgpulist.com/nebius-gpu-h200-sxm-1gpu-16vcpu-200gb/): Nebius gpu-h200-sxm / 1gpu-16vcpu-200gb supplies 1 NVIDIA H200 GPU.
- [gpu-h200-sxm / 8gpu-128vcpu-1600gb](https://cloudgpulist.com/nebius-gpu-h200-sxm-8gpu-128vcpu-1600gb/): Nebius gpu-h200-sxm / 8gpu-128vcpu-1600gb supplies 8 NVIDIA H200 GPUs.
- [gpu-h200x1-141gb](https://cloudgpulist.com/digitalocean-gpu-h200x1-141gb/): DigitalOcean gpu-h200x1-141gb allocates 1 NVIDIA H200 accelerator.
- [gpu-h200x8-1128gb](https://cloudgpulist.com/digitalocean-gpu-h200x8-1128gb/): DigitalOcean gpu-h200x8-1128gb allocates 8 NVIDIA H200 accelerators.
- [gpu-l40sx1-48gb](https://cloudgpulist.com/digitalocean-gpu-l40sx1-48gb/): DigitalOcean gpu-l40sx1-48gb allocates 1 NVIDIA L40S accelerator.
- [gpu-mi300x1-192gb](https://cloudgpulist.com/digitalocean-gpu-mi300x1-192gb/): DigitalOcean gpu-mi300x1-192gb allocates 1 AMD MI300X accelerator.
- [gpu-mi300x8-1536gb](https://cloudgpulist.com/digitalocean-gpu-mi300x8-1536gb/): DigitalOcean gpu-mi300x8-1536gb allocates 8 AMD MI300X accelerators.
- [GPU+ (M4000)](https://cloudgpulist.com/paperspace-gpu-m4000/): Paperspace GPU+ (M4000) allocates 1 NVIDIA M4000 accelerator.
- [GT4.20XLARGE474](https://cloudgpulist.com/tencent-gt4-20xlarge474/): Tencent GT4.20XLARGE474 allocates 4 NVIDIA A100 accelerators.
- [GT4.41XLARGE948](https://cloudgpulist.com/tencent-gt4-41xlarge948/): Tencent GT4.41XLARGE948 allocates 8 NVIDIA A100 accelerators.
- [GT4.4XLARGE96](https://cloudgpulist.com/tencent-gt4-4xlarge96/): Tencent GT4.4XLARGE96 allocates 1 NVIDIA A100 accelerator.
- [GT4.8XLARGE192](https://cloudgpulist.com/tencent-gt4-8xlarge192/): Tencent GT4.8XLARGE192 allocates 2 NVIDIA A100 accelerators.
- [gx3-16x80x1l4](https://cloudgpulist.com/ibm-cloud-gx3-16x80x1l4/): IBM Cloud gx3-16x80x1l4 allocates 1 NVIDIA L4 accelerator.
- [gx3-24x120x1l40s](https://cloudgpulist.com/ibm-cloud-gx3-24x120x1l40s/): IBM Cloud gx3-24x120x1l40s allocates 1 NVIDIA L40S accelerator.
- [gx3-32x160x2l4](https://cloudgpulist.com/ibm-cloud-gx3-32x160x2l4/): IBM Cloud gx3-32x160x2l4 allocates 2 NVIDIA L4 accelerators.
- [gx3-48x240x-2l40s](https://cloudgpulist.com/ibm-cloud-gx3-48x240x-2l40s/): IBM Cloud gx3-48x240x-2l40s allocates 2 NVIDIA L40S accelerators.
- [gx3-64x320x4l4](https://cloudgpulist.com/ibm-cloud-gx3-64x320x4l4/): IBM Cloud gx3-64x320x4l4 allocates 4 NVIDIA L4 accelerators.
- [gx3d-160x1792x8h100](https://cloudgpulist.com/ibm-cloud-gx3d-160x1792x8h100/): IBM Cloud gx3d-160x1792x8h100 allocates 8 NVIDIA H100 accelerators.
- [gx3d-160x1792x8h200](https://cloudgpulist.com/ibm-cloud-gx3d-160x1792x8h200/): IBM Cloud gx3d-160x1792x8h200 allocates 8 NVIDIA H200 accelerators.
- [gx3d-24x120x1a100p](https://cloudgpulist.com/ibm-cloud-gx3d-24x120x1a100p/): IBM Cloud gx3d-24x120x1a100p allocates 1 NVIDIA A100 PCIe accelerator.
- [gx3d-48x240x2a100p](https://cloudgpulist.com/ibm-cloud-gx3d-48x240x2a100p/): IBM Cloud gx3d-48x240x2a100p allocates 2 NVIDIA A100 PCIe accelerators.
- [H100](https://cloudgpulist.com/paperspace-h100/): Paperspace H100 allocates 1 NVIDIA H100 accelerator.
- [H100 NVL / 1 GPU Pod](https://cloudgpulist.com/runpod-h100-nvl-1-gpu-pod/): RunPod H100 NVL / 1 GPU Pod allocates 1 NVIDIA H100 NVL accelerator.
- [H100 PCIe / 1 GPU Pod](https://cloudgpulist.com/runpod-h100-pcie-1-gpu-pod/): RunPod H100 PCIe / 1 GPU Pod allocates 1 NVIDIA H100 PCIe accelerator.
- [H100 SXM / 1 GPU Pod](https://cloudgpulist.com/runpod-h100-sxm-1-gpu-pod/): RunPod H100 SXM / 1 GPU Pod allocates 1 NVIDIA H100 SXM accelerator.
- [H100-1-80G](https://cloudgpulist.com/scaleway-h100-1-80g/): Scaleway H100-1-80G allocates 1 NVIDIA H100-PCIe accelerator.
- [h100-1520](https://cloudgpulist.com/ovhcloud-h100-1520/): OVHcloud h100-1520 allocates 4 NVIDIA H100 accelerators.
- [H100-2-80G](https://cloudgpulist.com/scaleway-h100-2-80g/): Scaleway H100-2-80G allocates 2 NVIDIA H100-PCIe accelerators.
- [h100-380](https://cloudgpulist.com/ovhcloud-h100-380/): OVHcloud h100-380 allocates 1 NVIDIA H100 accelerator.
- [h100-760](https://cloudgpulist.com/ovhcloud-h100-760/): OVHcloud h100-760 allocates 2 NVIDIA H100 accelerators.
- [h100-80gb-sxm-ib.8x](https://cloudgpulist.com/crusoe-h100-80gb-sxm-ib-8x/): Crusoe’s H100 SXM instance combines eight GPUs with InfiniBand networking for multi-node workloads.
- [H100-SXM-2-80G](https://cloudgpulist.com/scaleway-h100-sxm-2-80g/): Scaleway H100-SXM-2-80G allocates 2 NVIDIA H100-SXM accelerators.
- [H100-SXM-4-80G](https://cloudgpulist.com/scaleway-h100-sxm-4-80g/): Scaleway H100-SXM-4-80G allocates 4 NVIDIA H100-SXM accelerators.
- [H100-SXM-8-80G](https://cloudgpulist.com/scaleway-h100-sxm-8-80g/): Scaleway H100-SXM-8-80G allocates 8 NVIDIA H100-SXM accelerators.
- [H100x8](https://cloudgpulist.com/paperspace-h100x8/): Paperspace H100x8 allocates 8 NVIDIA H100 accelerators.
- [H200 / 1 GPU Pod](https://cloudgpulist.com/runpod-h200-1-gpu-pod/): RunPod H200 / 1 GPU Pod allocates 1 NVIDIA H200 accelerator.
- [h200-141gb-sxm-ib.8x](https://cloudgpulist.com/crusoe-h200-141gb-sxm-ib-8x/): Crusoe h200-141gb-sxm-ib.8x allocates 8 NVIDIA H200 accelerators.
- [h200-1920](https://cloudgpulist.com/ovhcloud-h200-1920/): OVHcloud h200-1920 allocates 8 NVIDIA H200 accelerators.
- [h200-1920-eph](https://cloudgpulist.com/ovhcloud-h200-1920-eph/): OVHcloud h200-1920-eph allocates 8 NVIDIA H200 accelerators.
- [HGX B200 / 8 GPU node](https://cloudgpulist.com/together-hgx-b200-8-gpu-node/): Together HGX B200 / 8 GPU node allocates 8 NVIDIA B200 accelerators.
- [HGX H100 SXM / 8 GPU node](https://cloudgpulist.com/together-hgx-h100-sxm-8-gpu-node/): Together HGX H100 SXM / 8 GPU node allocates 8 NVIDIA H100 SXM accelerators.
- [HGX H200 / 8 GPU node](https://cloudgpulist.com/together-hgx-h200-8-gpu-node/): Together HGX H200 / 8 GPU node allocates 8 NVIDIA H200 accelerators.
- [L4 / 1 GPU Pod](https://cloudgpulist.com/runpod-l4-1-gpu-pod/): RunPod L4 / 1 GPU Pod allocates 1 NVIDIA L4 accelerator.
- [L4-1-24G](https://cloudgpulist.com/scaleway-l4-1-24g/): Scaleway L4-1-24G allocates 1 NVIDIA L4 accelerator.
- [l4-180](https://cloudgpulist.com/ovhcloud-l4-180/): OVHcloud l4-180 allocates 2 NVIDIA L4 accelerators.
- [L4-2-24G](https://cloudgpulist.com/scaleway-l4-2-24g/): Scaleway L4-2-24G allocates 2 NVIDIA L4 accelerators.
- [l4-360](https://cloudgpulist.com/ovhcloud-l4-360/): OVHcloud l4-360 allocates 4 NVIDIA L4 accelerators.
- [L4-4-24G](https://cloudgpulist.com/scaleway-l4-4-24g/): Scaleway L4-4-24G allocates 4 NVIDIA L4 accelerators.
- [L4-8-24G](https://cloudgpulist.com/scaleway-l4-8-24g/): Scaleway L4-8-24G allocates 8 NVIDIA L4 accelerators.
- [l4-90](https://cloudgpulist.com/ovhcloud-l4-90/): OVHcloud l4-90 allocates 1 NVIDIA L4 accelerator.
- [L40 / 1 GPU Pod](https://cloudgpulist.com/runpod-l40-1-gpu-pod/): RunPod L40 / 1 GPU Pod allocates 1 NVIDIA L40 accelerator.
- [L40S / 1 GPU Pod](https://cloudgpulist.com/runpod-l40s-1-gpu-pod/): RunPod L40S / 1 GPU Pod allocates 1 NVIDIA L40S accelerator.
- [L40S-1-48G](https://cloudgpulist.com/scaleway-l40s-1-48g/): Scaleway L40S-1-48G allocates 1 NVIDIA L40S accelerator.
- [l40s-180](https://cloudgpulist.com/ovhcloud-l40s-180/): OVHcloud l40s-180 allocates 2 NVIDIA L40S accelerators.
- [L40S-2-48G](https://cloudgpulist.com/scaleway-l40s-2-48g/): Scaleway L40S-2-48G allocates 2 NVIDIA L40S accelerators.
- [l40s-360](https://cloudgpulist.com/ovhcloud-l40s-360/): OVHcloud l40s-360 allocates 4 NVIDIA L40S accelerators.
- [L40S-4-48G](https://cloudgpulist.com/scaleway-l40s-4-48g/): Scaleway L40S-4-48G allocates 4 NVIDIA L40S accelerators.
- [l40s-48gb.10x](https://cloudgpulist.com/crusoe-l40s-48gb-10x/): Crusoe l40s-48gb.10x allocates 10 NVIDIA L40S accelerators.
- [l40s-48gb.1x](https://cloudgpulist.com/crusoe-l40s-48gb-1x/): Crusoe l40s-48gb.1x allocates 1 NVIDIA L40S accelerator.
- [l40s-48gb.2x](https://cloudgpulist.com/crusoe-l40s-48gb-2x/): Crusoe l40s-48gb.2x allocates 2 NVIDIA L40S accelerators.
- [l40s-48gb.4x](https://cloudgpulist.com/crusoe-l40s-48gb-4x/): Crusoe l40s-48gb.4x allocates 4 NVIDIA L40S accelerators.
- [l40s-48gb.8x](https://cloudgpulist.com/crusoe-l40s-48gb-8x/): Crusoe l40s-48gb.8x allocates 8 NVIDIA L40S accelerators.
- [L40S-8-48G](https://cloudgpulist.com/scaleway-l40s-8-48g/): Scaleway L40S-8-48G allocates 8 NVIDIA L40S accelerators.
- [l40s-90](https://cloudgpulist.com/ovhcloud-l40s-90/): OVHcloud l40s-90 allocates 1 NVIDIA L40S accelerator.
- [mi300x-192gb-ib.8x](https://cloudgpulist.com/crusoe-mi300x-192gb-ib-8x/): Crusoe mi300x-192gb-ib.8x allocates 8 AMD MI300X accelerators.
- [MI325X / 1 GPU Droplet](https://cloudgpulist.com/digitalocean-mi325x-1-gpu-droplet/): DigitalOcean MI325X / 1 GPU Droplet allocates 1 AMD MI325X accelerator.
- [MI325X / 8 GPU Droplet](https://cloudgpulist.com/digitalocean-mi325x-8-gpu-droplet/): DigitalOcean MI325X / 8 GPU Droplet allocates 8 AMD MI325X accelerators.
- [mi355x-288gb-roce.8x](https://cloudgpulist.com/crusoe-mi355x-288gb-roce-8x/): Crusoe mi355x-288gb-roce.8x allocates 8 AMD MI355X accelerators.
- [n3-A100-SXM4x8](https://cloudgpulist.com/hyperstack-n3-a100-sxm4x8/): Hyperstack n3-A100-SXM4x8 allocates 8 NVIDIA A100 SXM accelerators.
- [n3-A100x1](https://cloudgpulist.com/hyperstack-n3-a100x1/): Hyperstack n3-A100x1 allocates 1 NVIDIA A100 PCIe accelerator.
- [n3-A100x2](https://cloudgpulist.com/hyperstack-n3-a100x2/): Hyperstack n3-A100x2 allocates 2 NVIDIA A100 PCIe accelerators.
- [n3-A100x4](https://cloudgpulist.com/hyperstack-n3-a100x4/): Hyperstack n3-A100x4 allocates 4 NVIDIA A100 PCIe accelerators.
- [n3-A100x8](https://cloudgpulist.com/hyperstack-n3-a100x8/): Hyperstack n3-A100x8 allocates 8 NVIDIA A100 PCIe accelerators.
- [n3-A100x8-NVLink](https://cloudgpulist.com/hyperstack-n3-a100x8-nvlink/): Hyperstack n3-A100x8-NVLink allocates 8 NVIDIA A100 NVLink accelerators.
- [n3-B200-SXM6x8](https://cloudgpulist.com/hyperstack-n3-b200-sxm6x8/): Hyperstack n3-B200-SXM6x8 allocates 8 NVIDIA B200 accelerators.
- [n3-B300-SXM6x8](https://cloudgpulist.com/hyperstack-n3-b300-sxm6x8/): Hyperstack n3-B300-SXM6x8 allocates 8 NVIDIA B300 accelerators.
- [n3-H100-SXM5x8](https://cloudgpulist.com/hyperstack-n3-h100-sxm5x8/): Hyperstack n3-H100-SXM5x8 allocates 8 NVIDIA H100 SXM accelerators.
- [n3-H100x1](https://cloudgpulist.com/hyperstack-n3-h100x1/): Hyperstack n3-H100x1 allocates 1 NVIDIA H100 PCIe accelerator.
- [n3-H100x2](https://cloudgpulist.com/hyperstack-n3-h100x2/): Hyperstack n3-H100x2 allocates 2 NVIDIA H100 PCIe accelerators.
- [n3-H100x4](https://cloudgpulist.com/hyperstack-n3-h100x4/): Hyperstack n3-H100x4 allocates 4 NVIDIA H100 PCIe accelerators.
- [n3-H100x8](https://cloudgpulist.com/hyperstack-n3-h100x8/): Hyperstack n3-H100x8 allocates 8 NVIDIA H100 PCIe accelerators.
- [n3-H100x8-NVLink](https://cloudgpulist.com/hyperstack-n3-h100x8-nvlink/): Hyperstack n3-H100x8-NVLink allocates 8 NVIDIA H100 NVLink accelerators.
- [n3-H200-SXM5x8](https://cloudgpulist.com/hyperstack-n3-h200-sxm5x8/): Hyperstack n3-H200-SXM5x8 allocates 8 NVIDIA H200 SXM accelerators.
- [n3-L40x1](https://cloudgpulist.com/hyperstack-n3-l40x1/): Hyperstack n3-L40x1 allocates 1 NVIDIA L40 accelerator.
- [n3-L40x2](https://cloudgpulist.com/hyperstack-n3-l40x2/): Hyperstack n3-L40x2 allocates 2 NVIDIA L40 accelerators.
- [n3-L40x4](https://cloudgpulist.com/hyperstack-n3-l40x4/): Hyperstack n3-L40x4 allocates 4 NVIDIA L40 accelerators.
- [n3-L40x8](https://cloudgpulist.com/hyperstack-n3-l40x8/): Hyperstack n3-L40x8 allocates 8 NVIDIA L40 accelerators.
- [n3-RTX-A4000x1](https://cloudgpulist.com/hyperstack-n3-rtx-a4000x1/): Hyperstack n3-RTX-A4000x1 allocates 1 NVIDIA RTX A4000 accelerator.
- [n3-RTX-A4000x10](https://cloudgpulist.com/hyperstack-n3-rtx-a4000x10/): Hyperstack n3-RTX-A4000x10 allocates 10 NVIDIA RTX A4000 accelerators.
- [n3-RTX-A4000x2](https://cloudgpulist.com/hyperstack-n3-rtx-a4000x2/): Hyperstack n3-RTX-A4000x2 allocates 2 NVIDIA RTX A4000 accelerators.
- [n3-RTX-A4000x4](https://cloudgpulist.com/hyperstack-n3-rtx-a4000x4/): Hyperstack n3-RTX-A4000x4 allocates 4 NVIDIA RTX A4000 accelerators.
- [n3-RTX-A4000x8](https://cloudgpulist.com/hyperstack-n3-rtx-a4000x8/): Hyperstack n3-RTX-A4000x8 allocates 8 NVIDIA RTX A4000 accelerators.
- [n3-RTX-A6000x1](https://cloudgpulist.com/hyperstack-n3-rtx-a6000x1/): Hyperstack n3-RTX-A6000x1 allocates 1 NVIDIA RTX A6000 accelerator.
- [n3-RTX-A6000x2](https://cloudgpulist.com/hyperstack-n3-rtx-a6000x2/): Hyperstack n3-RTX-A6000x2 allocates 2 NVIDIA RTX A6000 accelerators.
- [n3-RTX-A6000x4](https://cloudgpulist.com/hyperstack-n3-rtx-a6000x4/): Hyperstack n3-RTX-A6000x4 allocates 4 NVIDIA RTX A6000 accelerators.
- [n3-RTX-A6000x8](https://cloudgpulist.com/hyperstack-n3-rtx-a6000x8/): Hyperstack n3-RTX-A6000x8 allocates 8 NVIDIA RTX A6000 accelerators.
- [n3-RTX-PRO6000-SEx1](https://cloudgpulist.com/hyperstack-n3-rtx-pro6000-sex1/): Hyperstack n3-RTX-PRO6000-SEx1 allocates 1 NVIDIA RTX PRO 6000 SE accelerator.
- [n3-RTX-PRO6000-SEx2](https://cloudgpulist.com/hyperstack-n3-rtx-pro6000-sex2/): Hyperstack n3-RTX-PRO6000-SEx2 allocates 2 NVIDIA RTX PRO 6000 SE accelerators.
- [n3-RTX-PRO6000-SEx4](https://cloudgpulist.com/hyperstack-n3-rtx-pro6000-sex4/): Hyperstack n3-RTX-PRO6000-SEx4 allocates 4 NVIDIA RTX PRO 6000 SE accelerators.
- [n3-RTX-PRO6000-SEx8](https://cloudgpulist.com/hyperstack-n3-rtx-pro6000-sex8/): Hyperstack n3-RTX-PRO6000-SEx8 allocates 8 NVIDIA RTX PRO 6000 SE accelerators.
- [p2vs.16xlarge.8](https://cloudgpulist.com/huawei-p2vs-16xlarge-8/): Huawei p2vs.16xlarge.8 allocates 8 NVIDIA V100 accelerators.
- [p2vs.2xlarge.8](https://cloudgpulist.com/huawei-p2vs-2xlarge-8/): Huawei p2vs.2xlarge.8 allocates 1 NVIDIA V100 accelerator.
- [p2vs.4xlarge.8](https://cloudgpulist.com/huawei-p2vs-4xlarge-8/): Huawei p2vs.4xlarge.8 allocates 2 NVIDIA V100 accelerators.
- [p2vs.8xlarge.8](https://cloudgpulist.com/huawei-p2vs-8xlarge-8/): Huawei p2vs.8xlarge.8 allocates 4 NVIDIA V100 accelerators.
- [P4000](https://cloudgpulist.com/paperspace-p4000/): Paperspace P4000 allocates 1 NVIDIA P4000 accelerator.
- [p4d.24xlarge](https://cloudgpulist.com/aws-p4d-24xlarge/): AWS p4d.24xlarge supplies 8 NVIDIA A100 GPUs.
- [p4de.24xlarge](https://cloudgpulist.com/aws-p4de-24xlarge/): AWS p4de.24xlarge supplies 8 NVIDIA A100 GPUs.
- [p5.48xlarge](https://cloudgpulist.com/aws-p5-48xlarge/): AWS p5.48xlarge supplies 8 NVIDIA H100 GPUs.
- [p5.4xlarge](https://cloudgpulist.com/aws-p5-4xlarge/): AWS p5.4xlarge supplies 1 NVIDIA H100 GPU.
- [P5000](https://cloudgpulist.com/paperspace-p5000/): Paperspace P5000 allocates 1 NVIDIA P5000 accelerator.
- [p5e.48xlarge](https://cloudgpulist.com/aws-p5e-48xlarge/): AWS p5e.48xlarge supplies 8 NVIDIA H200 GPUs.
- [p5en.48xlarge](https://cloudgpulist.com/aws-p5en-48xlarge/): AWS p5en.48xlarge supplies 8 NVIDIA H200 GPUs.
- [p6-b200.48xlarge](https://cloudgpulist.com/aws-p6-b200-48xlarge/): AWS p6-b200.48xlarge supplies 8 NVIDIA B200 GPUs.
- [p6-b300.48xlarge](https://cloudgpulist.com/aws-p6-b300-48xlarge/): AWS p6-b300.48xlarge allocates 8 NVIDIA B300 accelerators.
- [P6000](https://cloudgpulist.com/paperspace-p6000/): Paperspace P6000 allocates 1 NVIDIA P6000 accelerator.
- [p6e-gb200.36xlarge](https://cloudgpulist.com/aws-p6e-gb200-36xlarge/): AWS p6e-gb200.36xlarge allocates 4 NVIDIA B200 accelerators.
- [pi2.2xlarge.4](https://cloudgpulist.com/huawei-pi2-2xlarge-4/): Huawei pi2.2xlarge.4 allocates 1 NVIDIA T4 accelerator.
- [pi2.4xlarge.4](https://cloudgpulist.com/huawei-pi2-4xlarge-4/): Huawei pi2.4xlarge.4 allocates 2 NVIDIA T4 accelerators.
- [pi2.8xlarge.4](https://cloudgpulist.com/huawei-pi2-8xlarge-4/): Huawei pi2.8xlarge.4 allocates 4 NVIDIA T4 accelerators.
- [PNV4.14XLARGE232](https://cloudgpulist.com/tencent-pnv4-14xlarge232/): Tencent PNV4.14XLARGE232 allocates 2 NVIDIA A10 accelerators.
- [PNV4.28XLARGE466](https://cloudgpulist.com/tencent-pnv4-28xlarge466/): Tencent PNV4.28XLARGE466 allocates 4 NVIDIA A10 accelerators.
- [PNV4.56XLARGE932](https://cloudgpulist.com/tencent-pnv4-56xlarge932/): Tencent PNV4.56XLARGE932 allocates 8 NVIDIA A10 accelerators.
- [PNV4.7XLARGE116](https://cloudgpulist.com/tencent-pnv4-7xlarge116/): Tencent PNV4.7XLARGE116 allocates 1 NVIDIA A10 accelerator.
- [RTX 3090 / 1 GPU Pod](https://cloudgpulist.com/runpod-rtx-3090-1-gpu-pod/): RunPod RTX 3090 / 1 GPU Pod allocates 1 NVIDIA RTX 3090 accelerator.
- [RTX 4090 / 1 GPU Pod](https://cloudgpulist.com/runpod-rtx-4090-1-gpu-pod/): RunPod RTX 4090 / 1 GPU Pod allocates 1 NVIDIA RTX 4090 accelerator.
- [RTX 5090 / 1 GPU Pod](https://cloudgpulist.com/runpod-rtx-5090-1-gpu-pod/): RunPod RTX 5090 / 1 GPU Pod allocates 1 NVIDIA RTX 5090 accelerator.
- [RTX 6000 Ada / 1 GPU Pod](https://cloudgpulist.com/runpod-rtx-6000-ada-1-gpu-pod/): RunPod RTX 6000 Ada / 1 GPU Pod allocates 1 NVIDIA RTX 6000 Ada accelerator.
- [RTX A5000 / 1 GPU Pod](https://cloudgpulist.com/runpod-rtx-a5000-1-gpu-pod/): RunPod RTX A5000 / 1 GPU Pod allocates 1 NVIDIA RTX A5000 accelerator.
- [RTX A6000 / 1 GPU Pod](https://cloudgpulist.com/runpod-rtx-a6000-1-gpu-pod/): RunPod RTX A6000 / 1 GPU Pod allocates 1 NVIDIA RTX A6000 accelerator.
- [RTX Pro 6000 / 1 GPU Pod](https://cloudgpulist.com/runpod-rtx-pro-6000-1-gpu-pod/): RunPod RTX Pro 6000 / 1 GPU Pod allocates 1 NVIDIA RTX Pro 6000 accelerator.
- [RTX4000](https://cloudgpulist.com/paperspace-rtx4000/): Paperspace RTX4000 allocates 1 NVIDIA RTX4000 accelerator.
- [RTX5000](https://cloudgpulist.com/paperspace-rtx5000/): Paperspace RTX5000 allocates 1 NVIDIA RTX5000 accelerator.
- [rtx5000-28](https://cloudgpulist.com/ovhcloud-rtx5000-28/): OVHcloud rtx5000-28 allocates 1 NVIDIA Quadro-RTX5000 accelerator.
- [rtx5000-56](https://cloudgpulist.com/ovhcloud-rtx5000-56/): OVHcloud rtx5000-56 allocates 2 NVIDIA Quadro-RTX5000 accelerators.
- [rtx5000-84](https://cloudgpulist.com/ovhcloud-rtx5000-84/): OVHcloud rtx5000-84 allocates 3 NVIDIA Quadro-RTX5000 accelerators.
- [Standard_NC16as_T4_v3](https://cloudgpulist.com/azure-standard-nc16as-t4-v3/): Azure Standard_NC16as_T4_v3 allocates 1 NVIDIA T4 accelerator.
- [Standard_NC24ads_A100_v4](https://cloudgpulist.com/azure-standard-nc24ads-a100-v4/): Azure Standard_NC24ads_A100_v4 supplies 1 NVIDIA A100 PCIe GPU.
- [Standard_NC40ads_H100_v5](https://cloudgpulist.com/azure-standard-nc40ads-h100-v5/): Azure Standard_NC40ads_H100_v5 supplies 1 NVIDIA H100 NVL GPU.
- [Standard_NC48ads_A100_v4](https://cloudgpulist.com/azure-standard-nc48ads-a100-v4/): Azure Standard_NC48ads_A100_v4 supplies 2 NVIDIA A100 PCIe GPUs.
- [Standard_NC4as_T4_v3](https://cloudgpulist.com/azure-standard-nc4as-t4-v3/): Azure Standard_NC4as_T4_v3 allocates 1 NVIDIA T4 accelerator.
- [Standard_NC64as_T4_v3](https://cloudgpulist.com/azure-standard-nc64as-t4-v3/): Azure Standard_NC64as_T4_v3 allocates 4 NVIDIA T4 accelerators.
- [Standard_NC80adis_H100_v5](https://cloudgpulist.com/azure-standard-nc80adis-h100-v5/): Azure Standard_NC80adis_H100_v5 supplies 2 NVIDIA H100 NVL GPUs.
- [Standard_NC8as_T4_v3](https://cloudgpulist.com/azure-standard-nc8as-t4-v3/): Azure Standard_NC8as_T4_v3 allocates 1 NVIDIA T4 accelerator.
- [Standard_NC96ads_A100_v4](https://cloudgpulist.com/azure-standard-nc96ads-a100-v4/): Azure Standard_NC96ads_A100_v4 supplies 4 NVIDIA A100 PCIe GPUs.
- [Standard_ND128isr_NDR_GB200_v6](https://cloudgpulist.com/azure-standard-nd128isr-ndr-gb200-v6/): Azure Standard_ND128isr_NDR_GB200_v6 allocates 4 NVIDIA GB200 accelerators.
- [Standard_ND96amsr_A100_v4](https://cloudgpulist.com/azure-standard-nd96amsr-a100-v4/): Azure Standard_ND96amsr_A100_v4 supplies 8 NVIDIA A100 GPUs.
- [Standard_ND96isr_H100_v5](https://cloudgpulist.com/azure-standard-nd96isr-h100-v5/): Azure Standard_ND96isr_H100_v5 supplies 8 NVIDIA H100 GPUs.
- [Standard_ND96isr_H200_v5](https://cloudgpulist.com/azure-standard-nd96isr-h200-v5/): Azure Standard_ND96isr_H200_v5 allocates 8 NVIDIA H200 accelerators.
- [Standard_ND96isr_MI300X_v5](https://cloudgpulist.com/azure-standard-nd96isr-mi300x-v5/): Azure Standard_ND96isr_MI300X_v5 allocates 8 AMD MI300X accelerators.
- [t1-180](https://cloudgpulist.com/ovhcloud-t1-180/): OVHcloud t1-180 allocates 4 NVIDIA Tesla V100 accelerators.
- [t1-45](https://cloudgpulist.com/ovhcloud-t1-45/): OVHcloud t1-45 allocates 1 NVIDIA Tesla V100 accelerator.
- [t1-90](https://cloudgpulist.com/ovhcloud-t1-90/): OVHcloud t1-90 allocates 2 NVIDIA Tesla V100 accelerators.
- [t1-le-180](https://cloudgpulist.com/ovhcloud-t1-le-180/): OVHcloud t1-le-180 allocates 4 NVIDIA Tesla V100 accelerators.
- [t1-le-45](https://cloudgpulist.com/ovhcloud-t1-le-45/): OVHcloud t1-le-45 allocates 1 NVIDIA Tesla V100 accelerator.
- [t1-le-90](https://cloudgpulist.com/ovhcloud-t1-le-90/): OVHcloud t1-le-90 allocates 2 NVIDIA Tesla V100 accelerators.
- [t2-180](https://cloudgpulist.com/ovhcloud-t2-180/): OVHcloud t2-180 allocates 4 NVIDIA Tesla V100S accelerators.
- [t2-45](https://cloudgpulist.com/ovhcloud-t2-45/): OVHcloud t2-45 allocates 1 NVIDIA Tesla V100S accelerator.
- [t2-90](https://cloudgpulist.com/ovhcloud-t2-90/): OVHcloud t2-90 allocates 2 NVIDIA Tesla V100S accelerators.
- [t2-le-180](https://cloudgpulist.com/ovhcloud-t2-le-180/): OVHcloud t2-le-180 allocates 4 NVIDIA Tesla V100S accelerators.
- [t2-le-45](https://cloudgpulist.com/ovhcloud-t2-le-45/): OVHcloud t2-le-45 allocates 1 NVIDIA Tesla V100S accelerator.
- [t2-le-90](https://cloudgpulist.com/ovhcloud-t2-le-90/): OVHcloud t2-le-90 allocates 2 NVIDIA Tesla V100S accelerators.
- [V100](https://cloudgpulist.com/paperspace-v100/): Paperspace V100 allocates 1 NVIDIA V100 accelerator.
- [V100-32G](https://cloudgpulist.com/paperspace-v100-32g/): Paperspace V100-32G allocates 1 NVIDIA V100 accelerator.
- [vcg-l40s-16c-180g-48vram](https://cloudgpulist.com/vultr-vcg-l40s-16c-180g-48vram/): Vultr vcg-l40s-16c-180g-48vram allocates 1 NVIDIA L40S accelerator.
- [vcpu-12_memory-72g_nvidia-rtx-3090-3](https://cloudgpulist.com/genesis-cloud-vcpu-12-memory-72g-nvidia-rtx-3090-3/): Genesis Cloud vcpu-12_memory-72g_nvidia-rtx-3090-3 allocates 3 NVIDIA RTX 3090 accelerators.
- [vcpu-16_memory-96g_nvidia-rtx-3090-2](https://cloudgpulist.com/genesis-cloud-vcpu-16-memory-96g-nvidia-rtx-3090-2/): Genesis Cloud vcpu-16_memory-96g_nvidia-rtx-3090-2 allocates 2 NVIDIA RTX 3090 accelerators.
- [vcpu-16_memory-96g_nvidia-rtx-3090-4](https://cloudgpulist.com/genesis-cloud-vcpu-16-memory-96g-nvidia-rtx-3090-4/): Genesis Cloud vcpu-16_memory-96g_nvidia-rtx-3090-4 allocates 4 NVIDIA RTX 3090 accelerators.
- [vcpu-192_memory-1920g_nvidia-h100-sxm5-8](https://cloudgpulist.com/genesis-cloud-vcpu-192-memory-1920g-nvidia-h100-sxm5-8/): Genesis Cloud vcpu-192_memory-1920g_nvidia-h100-sxm5-8 allocates 8 NVIDIA H100 SXM accelerators.
- [vcpu-20_memory-120g_nvidia-rtx-3090-5](https://cloudgpulist.com/genesis-cloud-vcpu-20-memory-120g-nvidia-rtx-3090-5/): Genesis Cloud vcpu-20_memory-120g_nvidia-rtx-3090-5 allocates 5 NVIDIA RTX 3090 accelerators.
- [vcpu-24_memory-144g_nvidia-rtx-3090-3](https://cloudgpulist.com/genesis-cloud-vcpu-24-memory-144g-nvidia-rtx-3090-3/): Genesis Cloud vcpu-24_memory-144g_nvidia-rtx-3090-3 allocates 3 NVIDIA RTX 3090 accelerators.
- [vcpu-24_memory-144g_nvidia-rtx-3090-6](https://cloudgpulist.com/genesis-cloud-vcpu-24-memory-144g-nvidia-rtx-3090-6/): Genesis Cloud vcpu-24_memory-144g_nvidia-rtx-3090-6 allocates 6 NVIDIA RTX 3090 accelerators.
- [vcpu-24_memory-240g_nvidia-h100-sxm5-1](https://cloudgpulist.com/genesis-cloud-vcpu-24-memory-240g-nvidia-h100-sxm5-1/): Genesis Cloud vcpu-24_memory-240g_nvidia-h100-sxm5-1 allocates 1 NVIDIA H100 SXM accelerator.
- [vcpu-28_memory-168g_nvidia-rtx-3090-7](https://cloudgpulist.com/genesis-cloud-vcpu-28-memory-168g-nvidia-rtx-3090-7/): Genesis Cloud vcpu-28_memory-168g_nvidia-rtx-3090-7 allocates 7 NVIDIA RTX 3090 accelerators.
- [vcpu-32_memory-192g_nvidia-rtx-3090-4](https://cloudgpulist.com/genesis-cloud-vcpu-32-memory-192g-nvidia-rtx-3090-4/): Genesis Cloud vcpu-32_memory-192g_nvidia-rtx-3090-4 allocates 4 NVIDIA RTX 3090 accelerators.
- [vcpu-32_memory-192g_nvidia-rtx-3090-8](https://cloudgpulist.com/genesis-cloud-vcpu-32-memory-192g-nvidia-rtx-3090-8/): Genesis Cloud vcpu-32_memory-192g_nvidia-rtx-3090-8 allocates 8 NVIDIA RTX 3090 accelerators.
- [vcpu-4_memory-24g_nvidia-rtx-3090-1](https://cloudgpulist.com/genesis-cloud-vcpu-4-memory-24g-nvidia-rtx-3090-1/): Genesis Cloud vcpu-4_memory-24g_nvidia-rtx-3090-1 allocates 1 NVIDIA RTX 3090 accelerator.
- [vcpu-8_memory-48g_nvidia-rtx-3090-1](https://cloudgpulist.com/genesis-cloud-vcpu-8-memory-48g-nvidia-rtx-3090-1/): Genesis Cloud vcpu-8_memory-48g_nvidia-rtx-3090-1 allocates 1 NVIDIA RTX 3090 accelerator.
- [vcpu-8_memory-48g_nvidia-rtx-3090-2](https://cloudgpulist.com/genesis-cloud-vcpu-8-memory-48g-nvidia-rtx-3090-2/): Genesis Cloud vcpu-8_memory-48g_nvidia-rtx-3090-2 allocates 2 NVIDIA RTX 3090 accelerators.
- [VM.GPU.A10.1](https://cloudgpulist.com/oracle-vm-gpu-a10-1/): Oracle VM.GPU.A10.1 supplies 1 NVIDIA A10 GPU.
- [VM.GPU.A10.2](https://cloudgpulist.com/oracle-vm-gpu-a10-2/): Oracle VM.GPU.A10.2 allocates 2 NVIDIA A10 accelerators.

## Optional
- [All records in one file](https://cloudgpulist.com/llms-full.txt)
- [Sitemap](https://cloudgpulist.com/sitemap.xml)

## Configuration hubs
- [Single-GPU cloud instances](https://cloudgpulist.com/single-gpu/): Compare single-GPU cloud instances by model, device memory, host CPU and RAM, interconnect and scoped hourly price, with dated sources for every filled field.
- [Multi-GPU cloud instances](https://cloudgpulist.com/multi-gpu/): Compare multi-GPU cloud configurations by device count, per-GPU memory, host resources, interconnect and scoped hourly price, with original provider sources.
- [Cloud instances with at least 80 GB per GPU](https://cloudgpulist.com/large-memory/): Compare cloud configurations with at least 80 GB per GPU, including H100, H200, Blackwell and AMD options. Check memory units, host resources and price terms.
- [GPU instances from major public clouds](https://cloudgpulist.com/public-clouds/): Compare published GPU SKUs across major public clouds, with host specifications, device memory, regional scope and sourced hourly purchase terms side by side.
- [Specialist GPU clouds and dedicated servers](https://cloudgpulist.com/gpu-clouds/): Compare specialist GPU cloud presets, full GPU nodes and dedicated servers. Keep CPU resources, interconnect, source units and hourly purchase terms attached.