# GPUClouds

> Compare GPU clouds for training and inference.

Every GPU cloud provider on GPUClouds, grouped by city, with its facts. Each name links to the markdown copy of its page.

## Online, Global, United Arab Emirates

### [Core42 (Sovereign AI Cloud)](https://gpuclouds.fyi/ae/global/online/core42-sovereign-ai-cloud.md)

Core42's Sovereign AI Cloud offers NVIDIA H100, H200, B200, AMD MI300X, Cerebras and Qualcomm accelerators with InfiniBand networking, orchestrated through Kubernetes or Slurm, for training and inference.

- Page: https://gpuclouds.fyi/ae/global/online/core42-sovereign-ai-cloud
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, United Arab Emirates
- Website: <https://www.core42.ai/>
- Updated: 2026-10-29

## Online, Global, Australia

### [Firmus AI Cloud (Sustainable Metal Cloud)](https://gpuclouds.fyi/au/global/online/firmus-ai-cloud-sustainable-metal-cloud.md)

Asia-Pacific AI cloud from Firmus Technologies offering bare-metal NVIDIA GPU clusters and on-demand VMs (H200, H100, A100, L40S) with Slurm, built in its own energy-efficient AI factories in Australia and Singapore.

- Page: https://gpuclouds.fyi/au/global/online/firmus-ai-cloud-sustainable-metal-cloud
- Categories: GPU neoclouds, Bare-metal GPU clusters, Sovereign and regional GPU
- Address: Online, Global, Australia
- Website: <https://firmus.co/>
- Updated: 2026-11-07

### [IREN Cloud](https://gpuclouds.fyi/au/global/online/iren-cloud.md)

AI cloud from data center operator IREN offering bare-metal NVIDIA GPU clusters (H100, H200, B200, B300, GB300 NVL72) with InfiniBand, run in its own renewable-powered data centers.

- Page: https://gpuclouds.fyi/au/global/online/iren-cloud
- Categories: Bare-metal GPU clusters, GPU neoclouds
- Address: Online, Global, Australia
- Website: <https://iren.com/>
- Updated: 2026-11-07

### [Macquarie Cloud Services (Launch AI)](https://gpuclouds.fyi/au/global/online/macquarie-cloud-services-launch-ai.md)

Australian managed cloud provider offering Launch AI, a sovereign model-as-a-service platform on Red Hat OpenShift with shared or dedicated capacity, priced at fixed monthly capacity instead of tokens.

- Page: https://gpuclouds.fyi/au/global/online/macquarie-cloud-services-launch-ai
- Categories: Sovereign and regional GPU, Inference platforms
- Address: Online, Global, Australia
- Website: <https://www.macquariecloudservices.com/>
- Updated: 2026-10-29

### [ResetData](https://gpuclouds.fyi/au/global/online/resetdata.md)

Australian sovereign AI provider and NVIDIA Cloud Partner offering GPU-as-a-Service (bare metal, Kubernetes/OpenShift, clustered GPU job scheduling) and a serverless OpenAI-compatible inference API.

- Page: https://gpuclouds.fyi/au/global/online/resetdata
- Categories: Sovereign and regional GPU, Inference platforms
- Address: Online, Global, Australia
- Website: <https://resetdata.com.au/>
- Updated: 2026-11-07

### [Sharon AI](https://gpuclouds.fyi/au/global/online/sharon-ai.md)

Australian sovereign AI infrastructure provider offering on-demand and dedicated GPU cloud (NVIDIA H200, H100 NVL, L40S, A40; AMD MI300X) via bare metal, VMs, Docker and Kubernetes from Melbourne and Sydney data centres.

- Page: https://gpuclouds.fyi/au/global/online/sharon-ai
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Australia
- Website: <https://sharonai.com/>
- Updated: 2026-10-29

## Online, Global, Brazil

### [Magalu Cloud AI Hub](https://gpuclouds.fyi/br/global/online/magalu-cloud-ai-hub.md)

Provides hosted AI model inference through an OpenAI-compatible API, including language models, embeddings, rerankers and speech recognition, billed by token consumption.

- Page: https://gpuclouds.fyi/br/global/online/magalu-cloud-ai-hub
- Categories: Inference platforms, Sovereign and regional GPU
- Address: Online, Global, Brazil
- Website: <https://magalu.cloud/>
- Updated: 2026-11-07

## Online, Global, Canada

### [5C (Hypertec Cloud)](https://gpuclouds.fyi/ca/global/online/5c-hypertec-cloud.md)

AI infrastructure company formed when Hypertec Cloud acquired 5C Data Centers; offers bare-metal-as-a-service and managed customer-owned clusters on NVIDIA Blackwell, AMD Instinct and Intel Gaudi hardware.

- Page: https://gpuclouds.fyi/ca/global/online/5c-hypertec-cloud
- Categories: Bare-metal GPU clusters, GPU neoclouds
- Address: Online, Global, Canada
- Website: <https://5c.ai/>
- Updated: 2026-10-29

### [Denvr Dataworks](https://gpuclouds.fyi/ca/global/online/denvr-dataworks.md)

AI cloud offering on-demand NVIDIA GPU instances (H100, GH200, A100, A40) and Intel Gaudi 2 billed by the minute, plus reserved clusters (H200, B200, GB200, Gaudi 3, MI300X) and AI inference.

- Page: https://gpuclouds.fyi/ca/global/online/denvr-dataworks
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, Canada
- Website: <https://www.denvr.com/>
- Updated: 2026-10-29

### [Swan Chain](https://gpuclouds.fyi/ca/global/online/swan-chain.md)

Decentralized AI inference cloud offering one OpenAI-compatible API for frontier and open-source models, served by independent GPU providers and paid per token from prepaid credit.

- Page: https://gpuclouds.fyi/ca/global/online/swan-chain
- Categories: Decentralized GPU, Inference platforms
- Address: Online, Global, Canada
- Website: <https://swanchain.io/>
- Updated: 2026-10-29

## Online, Global, Switzerland

### [Exoscale](https://gpuclouds.fyi/ch/global/online/exoscale.md)

European cloud offering NVIDIA GPU instances with direct GPU passthrough, dedicated hypervisor options, per-second billing and managed Kubernetes integration for AI, HPC and rendering.

- Page: https://gpuclouds.fyi/ch/global/online/exoscale
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Switzerland
- Website: <https://www.exoscale.com/>
- Updated: 2026-11-07

### [Golem Network](https://gpuclouds.fyi/ch/global/online/golem-network.md)

Open-source decentralized platform where requestors rent computing power, including GPU providers for AI models, from peers, paying in the GLM token via the Yagna service, JS/Python SDKs and Ray interface.

- Page: https://gpuclouds.fyi/ch/global/online/golem-network
- Categories: Decentralized GPU
- Address: Online, Global, Switzerland
- Website: <https://golem.network/>
- Updated: 2026-10-29

### [Infomaniak](https://gpuclouds.fyi/ch/global/online/infomaniak.md)

Swiss OpenStack public cloud offering dedicated NVIDIA GPU instances, managed Kubernetes, storage and networking for AI and other compute workloads.

- Page: https://gpuclouds.fyi/ch/global/online/infomaniak
- Categories: Sovereign and regional GPU
- Address: Online, Global, Switzerland
- Website: <https://www.infomaniak.com/>
- Updated: 2026-10-29

## Online, Global, China

### [AutoDL](https://gpuclouds.fyi/cn/global/online/autodl.md)

AutoDL is a Chinese GPU rental cloud offering exclusive-GPU container instances (RTX 4090, A100, A800, H800, H20, L20 and more) with pay-as-you-go billing to the second or daily/weekly/monthly rental.

- Page: https://gpuclouds.fyi/cn/global/online/autodl
- Categories: GPU marketplaces, GPU neoclouds
- Address: Online, Global, China
- Website: <https://www.autodl.com/>
- Updated: 2026-10-29

### [Baidu AI Cloud (GPU Cloud Server)](https://gpuclouds.fyi/cn/global/online/baidu-ai-cloud-gpu-cloud-server.md)

Baidu AI Cloud GPU cloud servers provide NVIDIA A100, A10, L20, V100 and T4 instance families for AI training, inference, HPC and rendering, offered as subscription, pay-as-you-go or preemptible instances.

- Page: https://gpuclouds.fyi/cn/global/online/baidu-ai-cloud-gpu-cloud-server
- Categories: Hyperscalers
- Address: Online, Global, China
- Website: <https://cloud.baidu.com/>
- Updated: 2026-10-29

### [Huawei Cloud (GPU Accelerated Cloud Server)](https://gpuclouds.fyi/cn/global/online/huawei-cloud-gpu-accelerated-cloud-server.md)

Huawei Cloud GPU Accelerated Cloud Server (GACS) offers GPU and Ascend-based ECS instances for AI, scientific computing and graphics workstations, with pay-per-use, yearly/monthly and spot billing modes.

- Page: https://gpuclouds.fyi/cn/global/online/huawei-cloud-gpu-accelerated-cloud-server
- Categories: Hyperscalers
- Address: Online, Global, China
- Website: <https://www.huaweicloud.com/intl/en-us/>
- Updated: 2026-11-07

### [Matpool](https://gpuclouds.fyi/cn/global/online/matpool.md)

Chinese AI cloud provider (矩池云) offering GPU and CPU cloud servers billed by time or daily/weekly/monthly, cloud storage, offline jobs, a large-model API, team and education plans, and private cloud.

- Page: https://gpuclouds.fyi/cn/global/online/matpool
- Categories: GPU neoclouds, Sovereign and regional GPU
- Address: Online, Global, China
- Website: <https://matpool.com/>
- Updated: 2026-11-07

### [UCloud (GPU Cloud Server)](https://gpuclouds.fyi/cn/global/online/ucloud-gpu-cloud-server.md)

UCloud offers GPU cloud servers and 8-GPU bare-metal servers with RTX 40-series, P40, T4, V100, A100 and A800 cards, RDMA networking, preinstalled AI images, and hourly or monthly billing.

- Page: https://gpuclouds.fyi/cn/global/online/ucloud-gpu-cloud-server
- Categories: Hyperscalers, Bare-metal GPU clusters
- Address: Online, Global, China
- Website: <https://www.ucloud.cn/>
- Updated: 2026-10-29

### [Volcengine (GPU Cloud Server)](https://gpuclouds.fyi/cn/global/online/volcengine-gpu-cloud-server.md)

Volcengine, ByteDance's cloud platform in China, offers GPU cloud server instance families with NVIDIA H20, A100, A30, A10, V100 and T4 GPUs for AI training, inference and rendering.

- Page: https://gpuclouds.fyi/cn/global/online/volcengine-gpu-cloud-server
- Categories: Hyperscalers
- Address: Online, Global, China
- Website: <https://www.volcengine.com/>
- Updated: 2026-10-29

## Online, Global, Germany

### [Contabo](https://gpuclouds.fyi/de/global/online/contabo.md)

VPS and bare-metal hosting provider offering a GPU VPS with a dedicated NVIDIA RTX PRO 6000 (96 GB) via full PCIe passthrough, at a flat monthly rate, in Europe or US Central; multi-GPU servers via vshosting.

- Page: https://gpuclouds.fyi/de/global/online/contabo
- Categories: Bare-metal GPU clusters
- Address: Online, Global, Germany
- Website: <https://contabo.com/>
- Updated: 2026-11-07

### [Deutsche Telekom Industrial AI Cloud](https://gpuclouds.fyi/de/global/online/deutsche-telekom-industrial-ai-cloud.md)

Deutsche Telekom and T-Systems provide on-demand GPU capacity from Munich for AI model training, inference and industrial simulations on a sovereign European platform.

- Page: https://gpuclouds.fyi/de/global/online/deutsche-telekom-industrial-ai-cloud
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Germany
- Website: <https://www.telekom.com/>
- Updated: 2026-10-29

### [Hetzner](https://gpuclouds.fyi/de/global/online/hetzner.md)

German hosting provider offering GEX-line dedicated GPU servers with NVIDIA RTX PRO 4000 and RTX PRO 6000 Blackwell GPUs for AI inference and training, billed monthly or hourly.

- Page: https://gpuclouds.fyi/de/global/online/hetzner
- Categories: Bare-metal GPU clusters
- Address: Online, Global, Germany
- Website: <https://www.hetzner.com/>
- Updated: 2026-10-29

### [IONOS](https://gpuclouds.fyi/de/global/online/ionos.md)

Cloud offering H200 GPU virtual machines hosted in Germany and dedicated GPU servers for AI training, inference, analytics and rendering, with usage billed by the minute.

- Page: https://gpuclouds.fyi/de/global/online/ionos
- Categories: Sovereign and regional GPU, Bare-metal GPU clusters, GPU neoclouds
- Address: Online, Global, Germany
- Website: <https://www.ionos.com/>
- Updated: 2026-11-07

### [Lyceum](https://gpuclouds.fyi/de/global/online/lyceum.md)

Offers GPU VMs, reserved clusters, serverless training and model inference in Europe, with per-second GPU billing and on-demand, spot and reserved capacity.

- Page: https://gpuclouds.fyi/de/global/online/lyceum
- Categories: GPU neoclouds, Inference platforms, Serverless GPU
- Address: Online, Global, Germany
- Website: <https://lyceum.technology/>
- Updated: 2026-11-07

### [Polarise](https://gpuclouds.fyi/de/global/online/polarise.md)

Offers dedicated NVIDIA B300 and GB300 GPU infrastructure in European data centers for AI training, fine-tuning and inference, with bare-metal and GPU-as-a-Service options.

- Page: https://gpuclouds.fyi/de/global/online/polarise
- Categories: Bare-metal GPU clusters, Sovereign and regional GPU
- Address: Online, Global, Germany
- Website: <https://polarise.eu/>
- Updated: 2026-10-29

### [STACKIT](https://gpuclouds.fyi/de/global/online/stackit.md)

European cloud from Schwarz Group offering GPU virtual machines with one to eight NVIDIA GPUs for AI, machine learning and rendering, plus managed Kubernetes and model serving.

- Page: https://gpuclouds.fyi/de/global/online/stackit
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Germany
- Website: <https://stackit.com/>
- Updated: 2026-10-29

### [T Cloud Public (formerly Open Telekom Cloud)](https://gpuclouds.fyi/de/global/online/t-cloud-public-formerly-open-telekom-cloud.md)

European OpenStack cloud from T-Systems offering GPU-accelerated virtual machines for deep learning, inference, scientific computing and graphics, with usage-based and reserved pricing.

- Page: https://gpuclouds.fyi/de/global/online/t-cloud-public-formerly-open-telekom-cloud
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Germany
- Website: <https://www.t-cloud-public.com/>
- Updated: 2026-10-29

## Online, Global, Finland

### [UpCloud](https://gpuclouds.fyi/fi/global/online/upcloud.md)

European cloud offering on-demand, spot and reserved NVIDIA GPU servers in Helsinki, with dedicated GPU passthrough, hourly billing and integration with its cloud services.

- Page: https://gpuclouds.fyi/fi/global/online/upcloud
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Finland
- Website: <https://upcloud.com/>
- Updated: 2026-11-07

### [Verda (formerly DataCrunch)](https://gpuclouds.fyi/fi/global/online/verda-formerly-datacrunch.md)

European AI cloud (formerly DataCrunch) offering pay-as-you-go, spot and reserved NVIDIA GPU instances (GB300, B300, B200, H200, H100, A100, L40S), instant clusters with Slurm or Kubernetes, and serverless containers.

- Page: https://gpuclouds.fyi/fi/global/online/verda-formerly-datacrunch
- Categories: GPU neoclouds, Serverless GPU, Sovereign and regional GPU
- Address: Online, Global, Finland
- Website: <https://verda.com/>
- Updated: 2026-11-07

## Online, Global, France

### [Ikoula](https://gpuclouds.fyi/fr/global/online/ikoula.md)

Rents dedicated GPU servers hosted in France for AI, rendering, simulation and deep learning, with single- and dual-GPU configurations billed monthly.

- Page: https://gpuclouds.fyi/fr/global/online/ikoula
- Categories: Bare-metal GPU clusters, Sovereign and regional GPU
- Address: Online, Global, France
- Website: <https://www.ikoula.com/>
- Updated: 2026-11-07

### [Koyeb](https://gpuclouds.fyi/fr/global/online/koyeb.md)

Serverless platform that deploys apps and inference endpoints in autoscaling VMs on bare-metal servers, with GPUs from L4 and RTX A6000 to A100, H100, H200 and B200, scale-to-zero and per-second billing.

- Page: https://gpuclouds.fyi/fr/global/online/koyeb
- Categories: Serverless GPU, Inference platforms
- Address: Online, Global, France
- Website: <https://www.koyeb.com/>
- Updated: 2026-10-29

### [OVHcloud](https://gpuclouds.fyi/fr/global/online/ovhcloud.md)

Cloud provider offering NVIDIA GPU instances, managed Kubernetes, AI notebooks, model training, application deployment and serverless model APIs with usage-based billing.

- Page: https://gpuclouds.fyi/fr/global/online/ovhcloud
- Categories: Sovereign and regional GPU, GPU neoclouds, Inference platforms
- Address: Online, Global, France
- Website: <https://www.ovhcloud.com/>
- Updated: 2026-10-29

### [Sesterce](https://gpuclouds.fyi/fr/global/online/sesterce.md)

Dedicated bare-metal GPU clusters for AI workloads, with single-tenant hardware, networking, storage, root access and infrastructure operations through Sesterce OS.

- Page: https://gpuclouds.fyi/fr/global/online/sesterce
- Categories: Bare-metal GPU clusters, GPU neoclouds
- Address: Online, Global, France
- Website: <https://www.sesterce.com/>
- Updated: 2026-10-29

## Online, Global, United Kingdom

### [Cassava AI](https://gpuclouds.fyi/gb/global/online/cassava-ai.md)

Offers NVIDIA H200 GPU-as-a-Service with isolated MIG instances and on-demand hourly, reserved and interruptible spot capacity for AI workloads.

- Page: https://gpuclouds.fyi/gb/global/online/cassava-ai
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, United Kingdom
- Website: <https://www.cassava.ai/>
- Updated: 2026-10-29

### [Civo](https://gpuclouds.fyi/gb/global/online/civo.md)

Cloud provider offering NVIDIA GPUs (B200, H200, H100, A100, L40S) on its compute and Kubernetes services with hourly on-demand or committed pricing, plus private cloud options.

- Page: https://gpuclouds.fyi/gb/global/online/civo
- Categories: GPU neoclouds, Sovereign and regional GPU
- Address: Online, Global, United Kingdom
- Website: <https://www.civo.com/>
- Updated: 2026-11-07

### [CUDO Compute](https://gpuclouds.fyi/gb/global/online/cudo-compute.md)

Dedicated NVIDIA GPU clusters and managed AI infrastructure for training and inference, with deployment, networking, storage and operational support across multiple regions.

- Page: https://gpuclouds.fyi/gb/global/online/cudo-compute
- Categories: Bare-metal GPU clusters, GPU neoclouds
- Address: Online, Global, United Kingdom
- Website: <https://www.cudocompute.com/>
- Updated: 2026-10-29

### [Flux (FluxEdge)](https://gpuclouds.fyi/gb/global/online/flux-fluxedge.md)

Flux is a decentralized cloud of user-operated nodes; its FluxEdge product is a P2P compute marketplace connecting developers with hardware providers renting dedicated GPUs, with Kubernetes-based container orchestration.

- Page: https://gpuclouds.fyi/gb/global/online/flux-fluxedge
- Categories: Decentralized GPU, GPU marketplaces
- Address: Online, Global, United Kingdom
- Website: <https://runonflux.com/>
- Updated: 2026-11-07

### [Hyperstack](https://gpuclouds.fyi/gb/global/online/hyperstack.md)

GPU cloud by NexGen Cloud offering on-demand, reserved and spot NVIDIA GPU VMs (B300, B200, H200, H100, A100, L40, RTX Pro 6000) billed per minute, plus on-demand Kubernetes and AI Studio.

- Page: https://gpuclouds.fyi/gb/global/online/hyperstack
- Categories: GPU neoclouds
- Address: Online, Global, United Kingdom
- Website: <https://www.hyperstack.cloud/>
- Updated: 2026-11-07

### [NetMind AI](https://gpuclouds.fyi/gb/global/online/netmind-ai.md)

AI platform offering a unified API to 100+ models plus GPU compute (H100, A100, RTX 4090) on-demand or reserved, GPU clusters for training, and dedicated inference endpoints.

- Page: https://gpuclouds.fyi/gb/global/online/netmind-ai
- Categories: Inference platforms, Decentralized GPU, GPU neoclouds
- Address: Online, Global, United Kingdom
- Website: <https://www.netmind.ai/>
- Updated: 2026-11-07

### [Northflank](https://gpuclouds.fyi/gb/global/online/northflank.md)

Container deployment platform running workloads on Kubernetes in Northflank Cloud or the customer's own cloud, with on-demand NVIDIA GPUs (L4, A100, H100, H200, B200, RTX PRO 6000) billed by the second.

- Page: https://gpuclouds.fyi/gb/global/online/northflank
- Categories: Serverless GPU, GPU neoclouds
- Address: Online, Global, United Kingdom
- Website: <https://northflank.com/>
- Updated: 2026-11-07

### [Nscale](https://gpuclouds.fyi/gb/global/online/nscale.md)

AI cloud offering bare-metal and on-demand NVIDIA GPU compute, managed Kubernetes and Slurm, and inference endpoints, run from its own and partner data centers in Norway, the UK, Iceland, Portugal and the US.

- Page: https://gpuclouds.fyi/gb/global/online/nscale
- Categories: GPU neoclouds, Inference platforms, Sovereign and regional GPU
- Address: Online, Global, United Kingdom
- Website: <https://www.nscale.com/>
- Updated: 2026-10-29

### [OctaSpace](https://gpuclouds.fyi/gb/global/online/octaspace.md)

Decentralized GPU cloud where users rent GPU nodes from a global network with live hourly pricing for AI, rendering (OctaRender) and app workloads, accessed via SSH or WebSSH; also runs a decentralized VPN.

- Page: https://gpuclouds.fyi/gb/global/online/octaspace
- Categories: Decentralized GPU, GPU marketplaces
- Address: Online, Global, United Kingdom
- Website: <https://octa.space/>
- Updated: 2026-11-07

### [Radiant (formerly Ori)](https://gpuclouds.fyi/gb/global/online/radiant-formerly-ori.md)

AI cloud formed from the merger of Ori Industries and Brookfield AI infrastructure; Ori's GPU cloud continues as Radiant AI Cloud with GPU VMs, bare metal, serverless Kubernetes and supercomputer clusters.

- Page: https://gpuclouds.fyi/gb/global/online/radiant-formerly-ori
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United Kingdom
- Website: <https://radiant.co/>
- Updated: 2026-10-29

### [Runware](https://gpuclouds.fyi/gb/global/online/runware.md)

AI cloud offering model APIs, autoscaling serverless GPU workloads and reserved bare-metal clusters on its modular Sonic Pod infrastructure, with GPU usage metered per second.

- Page: https://gpuclouds.fyi/gb/global/online/runware
- Categories: Inference platforms, Bare-metal GPU clusters, Serverless GPU
- Address: Online, Global, United Kingdom
- Website: <https://runware.ai/>
- Updated: 2026-11-07

## Online, Global, Indonesia

### [Indosat / Lintasarta Cloudeka GPU Merdeka](https://gpuclouds.fyi/id/global/online/indosat-lintasarta-cloudeka-gpu-merdeka.md)

Indonesian GPU cloud offering NVIDIA H100 and L40S instances for AI, machine learning and rendering, with hourly or monthly pricing and Kubernetes support.

- Page: https://gpuclouds.fyi/id/global/online/indosat-lintasarta-cloudeka-gpu-merdeka
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Indonesia
- Website: <https://www.cloudeka.id/>
- Updated: 2026-10-29

### [Lintasarta Deka GPU (Cloudeka)](https://gpuclouds.fyi/id/global/online/lintasarta-deka-gpu-cloudeka.md)

Lintasarta's Deka GPU, part of Cloudeka, is an Indonesian cloud GPU service with NVIDIA H100 SXM and L40S, including vGPU and MIG slices, priced hourly in IDR with a 10-day minimum period.

- Page: https://gpuclouds.fyi/id/global/online/lintasarta-deka-gpu-cloudeka
- Categories: Sovereign and regional GPU
- Address: Online, Global, Indonesia
- Website: <https://www.lintasarta.net/>
- Updated: 2026-10-29

## Online, Global, India

### [Cyfuture AI](https://gpuclouds.fyi/in/global/online/cyfuture-ai.md)

Noida-based AI infrastructure provider renting NVIDIA B300, H100, A100, L40S, RTX PRO 6000 and V100 GPUs hourly or reserved, with GPU clusters, managed Kubernetes, fine-tuning and serverless inference.

- Page: https://gpuclouds.fyi/in/global/online/cyfuture-ai
- Categories: GPU neoclouds, Inference platforms, Sovereign and regional GPU
- Address: Online, Global, India
- Website: <https://cyfuture.ai/>
- Updated: 2026-11-07

### [E2E Networks](https://gpuclouds.fyi/in/global/online/e2e-networks.md)

India-based, MeitY-empaneled GPU cloud offering on-demand NVIDIA B200, H200, H100, A100 and other GPU instances, training clusters, inference endpoints and managed Kubernetes.

- Page: https://gpuclouds.fyi/in/global/online/e2e-networks
- Categories: GPU neoclouds, Sovereign and regional GPU
- Address: Online, Global, India
- Website: <https://www.e2enetworks.com/>
- Updated: 2026-11-07

### [GPUNET](https://gpuclouds.fyi/in/global/online/gpunet.md)

GPU aggregator bringing bare-metal GPUs, datacenters and colocation from multiple partners under one contract, with on-demand GPU rental and reserved clusters on 3 to 24 month terms.

- Page: https://gpuclouds.fyi/in/global/online/gpunet
- Categories: GPU marketplaces, Bare-metal GPU clusters, Decentralized GPU
- Address: Online, Global, India
- Website: <https://www.gpu.net/>
- Updated: 2026-10-29

### [Jarvislabs](https://gpuclouds.fyi/in/global/online/jarvislabs.md)

GPU cloud offering on-demand and spot NVIDIA GPU instances (H200, H100, RTX Pro 6000, A100, A30, L4) billed per minute, instant H200 clusters, serverless and managed model endpoints, and a token API.

- Page: https://gpuclouds.fyi/in/global/online/jarvislabs
- Categories: GPU neoclouds, Inference platforms, Serverless GPU
- Address: Online, Global, India
- Website: <https://jarvislabs.ai/>
- Updated: 2026-11-07

### [Krutrim Cloud (Ola Krutrim)](https://gpuclouds.fyi/in/global/online/krutrim-cloud-ola-krutrim.md)

Ola Group's India-hosted AI cloud offering NVIDIA H100 and A100 GPU instances, Kubernetes-managed fractional GPU pods, sandboxes, CPU instances and managed Kubernetes, billed hourly with reserved options.

- Page: https://gpuclouds.fyi/in/global/online/krutrim-cloud-ola-krutrim
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, India
- Website: <https://www.olakrutrim.com/>
- Updated: 2026-11-07

### [NeevCloud](https://gpuclouds.fyi/in/global/online/neevcloud.md)

Indian AI cloud offering on-demand and reserved NVIDIA H100, A100, L40S and RTX PRO 6000 GPUs as Kubernetes-managed containers or VMs in India datacenters, plus model APIs and serverless inference.

- Page: https://gpuclouds.fyi/in/global/online/neevcloud
- Categories: GPU neoclouds, Inference platforms, Sovereign and regional GPU
- Address: Online, Global, India
- Website: <https://neevcloud.com/>
- Updated: 2026-10-29

### [Tata Communications Vayu AI Cloud (GPU as a Service)](https://gpuclouds.fyi/in/global/online/tata-communications-vayu-ai-cloud-gpu-as-a-service.md)

Tata Communications' Vayu AI Cloud offers GPU as a Service with bare-metal 8x H100 SXM and H200 nodes on InfiniBand plus H200 and L40S GPU VMs, priced hourly in INR, alongside managed Kubernetes.

- Page: https://gpuclouds.fyi/in/global/online/tata-communications-vayu-ai-cloud-gpu-as-a-service
- Categories: Sovereign and regional GPU, Bare-metal GPU clusters
- Address: Online, Global, India
- Website: <https://www.tatacommunications.com/>
- Updated: 2026-10-29

### [Yotta Shakti Cloud](https://gpuclouds.fyi/in/global/online/yotta-shakti-cloud.md)

Sovereign Indian AI cloud from Yotta Data Services offering NVIDIA H100, L40S, B200 and B300 GPU VMs, bare metal, SLURM/Kubernetes clusters, serverless GPUs and AI endpoints hosted in India.

- Page: https://gpuclouds.fyi/in/global/online/yotta-shakti-cloud
- Categories: Sovereign and regional GPU, Bare-metal GPU clusters, GPU neoclouds
- Address: Online, Global, India
- Website: <https://shakticloud.ai/>
- Updated: 2026-10-29

## Online, Global, Italy

### [Aruba Cloud](https://gpuclouds.fyi/it/global/online/aruba-cloud.md)

Italian cloud provider (Aruba S.p.A.) offering GPU on Demand: dedicated NVIDIA L40S and RTX 4080 Super GPU servers federated into a customer's Kubernetes cluster, billed monthly, with data kept in Europe.

- Page: https://gpuclouds.fyi/it/global/online/aruba-cloud
- Categories: Sovereign and regional GPU, Bare-metal GPU clusters
- Address: Online, Global, Italy
- Website: <https://www.arubacloud.com/>
- Updated: 2026-10-29

### [Seeweb](https://gpuclouds.fyi/it/global/online/seeweb.md)

Italian cloud provider offering on-demand and spot GPU servers for AI, machine learning and HPC, with hourly billing, Kubernetes integration and infrastructure-as-code support.

- Page: https://gpuclouds.fyi/it/global/online/seeweb
- Categories: Sovereign and regional GPU, GPU neoclouds, Serverless GPU
- Address: Online, Global, Italy
- Website: <https://www.seeweb.it/>
- Updated: 2026-11-07

## Online, Global, Japan

### [GMO GPU Cloud](https://gpuclouds.fyi/jp/global/online/gmo-gpu-cloud.md)

GMO Internet's GPU cloud offers NVIDIA HGX B300 and H200 servers from Japanese data centers as a managed HPC cluster service with Slurm and as a bare-metal service, with monthly quoted pricing and free data transfer.

- Page: https://gpuclouds.fyi/jp/global/online/gmo-gpu-cloud
- Categories: Sovereign and regional GPU, Bare-metal GPU clusters
- Address: Online, Global, Japan
- Website: <https://gpucloud.gmo/>
- Updated: 2026-10-29

### [Highreso GPUSOROBAN](https://gpuclouds.fyi/jp/global/online/highreso-gpusoroban.md)

Highreso's GPUSOROBAN is a Japan-based GPU cloud offering A4000 and A100 instances billed hourly or monthly in JPY, plus H200 AI supercomputer cloud and B200 compute clusters, with no data transfer fees.

- Page: https://gpuclouds.fyi/jp/global/online/highreso-gpusoroban
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Japan
- Website: <https://highreso.jp/>
- Updated: 2026-10-29

### [Sakura Internet (Koukaryoku GPU cloud)](https://gpuclouds.fyi/jp/global/online/sakura-internet-koukaryoku-gpu-cloud.md)

Sakura Internet's Koukaryoku series provides GPU cloud from Japanese data centers: DOK container GPU billed per second, VRT GPU VMs billed hourly, and PHY bare-metal servers with 8x H100, H200 or B200.

- Page: https://gpuclouds.fyi/jp/global/online/sakura-internet-koukaryoku-gpu-cloud
- Categories: Sovereign and regional GPU, Bare-metal GPU clusters
- Address: Online, Global, Japan
- Website: <https://www.sakura.ad.jp/>
- Updated: 2026-10-29

## Online, Global, St. Kitts & Nevis

### [Lium](https://gpuclouds.fyi/kn/global/online/lium.md)

GPU rental marketplace where providers list machines that the platform checks and scores, and renters launch GPU pods from RTX 5090 to B300 billed per second via console, CLI, SDK or API.

- Page: https://gpuclouds.fyi/kn/global/online/lium
- Categories: GPU marketplaces, Decentralized GPU
- Address: Online, Global, St. Kitts & Nevis
- Website: <https://lium.io/>
- Updated: 2026-11-07

## Online, Global, South Korea

### [Elice Cloud](https://gpuclouds.fyi/kr/global/online/elice-cloud.md)

Elice Cloud offers Runbox on-demand and reserved GPU instances including H100, B200 GPU clusters available for reservation in Korea, serverless ML APIs and DataHub storage, billed on actual instance usage time.

- Page: https://gpuclouds.fyi/kr/global/online/elice-cloud
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, South Korea
- Website: <https://elice.io/>
- Updated: 2026-11-07

### [Lablup Backend.AI Cloud](https://gpuclouds.fyi/kr/global/online/lablup-backend-ai-cloud.md)

Lablup's Backend.AI is a GPU orchestration and fractional GPU virtualization platform, available on-premises or as a cloud service (trial and enterprise) running on Lablup's resources for ML training and inference.

- Page: https://gpuclouds.fyi/kr/global/online/lablup-backend-ai-cloud
- Categories: GPU neoclouds
- Address: Online, Global, South Korea
- Website: <https://www.backend.ai/>
- Updated: 2026-10-29

### [NAVER Cloud Platform (GPU Server)](https://gpuclouds.fyi/kr/global/online/naver-cloud-platform-gpu-server.md)

NAVER Cloud Platform offers GPU Server as KVM/XEN virtual machines and bare metal servers with NVIDIA H200, H100, A100, L40S, L4, V100 and T4 GPUs, priced in KRW with hourly pay-as-you-go and monthly rates.

- Page: https://gpuclouds.fyi/kr/global/online/naver-cloud-platform-gpu-server
- Categories: Sovereign and regional GPU, Hyperscalers
- Address: Online, Global, South Korea
- Website: <https://www.ncloud.com/>
- Updated: 2026-10-29

### [Samsung Cloud Platform](https://gpuclouds.fyi/kr/global/online/samsung-cloud-platform.md)

Samsung SDS offers on-demand GPU-as-a-Service for AI development and large language model training, including NVIDIA H100 GPU clusters.

- Page: https://gpuclouds.fyi/kr/global/online/samsung-cloud-platform
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, South Korea
- Website: <https://www.samsungsds.com/>
- Updated: 2026-10-29

## Online, Global, Cayman Islands

### [Akash Network](https://gpuclouds.fyi/ky/global/online/akash-network.md)

Decentralized compute marketplace where independent providers bid to host containerized workloads on GPUs and CPUs; also offers AkashML, OpenAI-compatible inference APIs for open-source models.

- Page: https://gpuclouds.fyi/ky/global/online/akash-network
- Categories: Decentralized GPU, GPU marketplaces, Inference platforms
- Address: Online, Global, Cayman Islands
- Website: <https://akash.network/>
- Updated: 2026-10-29

### [Livepeer](https://gpuclouds.fyi/ky/global/online/livepeer.md)

Open, decentralized inference network for AI video workloads; independent GPU providers connect GPUs and set prices, and Livepeer Agent routes image, video, audio and 3D generation jobs to them.

- Page: https://gpuclouds.fyi/ky/global/online/livepeer
- Categories: Decentralized GPU, Inference platforms
- Address: Online, Global, Cayman Islands
- Website: <https://livepeer.org/>
- Updated: 2026-10-29

## Online, Global, Lithuania

### [Cherry Servers](https://gpuclouds.fyi/lt/global/online/cherry-servers.md)

Lithuania-based provider of bare-metal dedicated servers with optional NVIDIA GPUs (A100, A40, A16, A10, A2 and others) for AI, ML and data processing, with hourly, monthly or yearly billing.

- Page: https://gpuclouds.fyi/lt/global/online/cherry-servers
- Categories: Bare-metal GPU clusters
- Address: Online, Global, Lithuania
- Website: <https://www.cherryservers.com/>
- Updated: 2026-11-07

## Online, Global, Luxembourg

### [Gcore](https://gpuclouds.fyi/lu/global/online/gcore.md)

Cloud offering NVIDIA GPU virtual machines and bare-metal servers for AI training, inference and HPC, with multi-GPU clusters, InfiniBand and Kubernetes support.

- Page: https://gpuclouds.fyi/lu/global/online/gcore
- Categories: GPU neoclouds, Bare-metal GPU clusters, Inference platforms
- Address: Online, Global, Luxembourg
- Website: <https://gcore.com/>
- Updated: 2026-11-07

## Online, Global, Malaysia

### [YTL AI Cloud](https://gpuclouds.fyi/my/global/online/ytl-ai-cloud.md)

Provides AI infrastructure as a service for model training and inference, with NVIDIA GB200 GPU clusters based in Johor, Malaysia.

- Page: https://gpuclouds.fyi/my/global/online/ytl-ai-cloud
- Categories: Sovereign and regional GPU, Bare-metal GPU clusters
- Address: Online, Global, Malaysia
- Website: <https://www.ytlaicloud.com/>
- Updated: 2026-10-29

## Online, Global, Netherlands

### [HOSTKEY](https://gpuclouds.fyi/nl/global/online/hostkey.md)

Hosting provider renting dedicated GPU servers and GPU VPS with NVIDIA (H100, A100, RTX PRO 6000, RTX 4090/5090, A-series) and AMD Radeon GPUs, with hourly or monthly billing in European and US data centers.

- Page: https://gpuclouds.fyi/nl/global/online/hostkey
- Categories: Bare-metal GPU clusters
- Address: Online, Global, Netherlands
- Website: <https://hostkey.com/>
- Updated: 2026-11-07

### [Leafcloud](https://gpuclouds.fyi/nl/global/online/leafcloud.md)

Dutch OpenStack cloud offering NVIDIA GPU virtual machines and managed Kubernetes in Amsterdam, with server heat reused to warm buildings and pay-as-you-go or committed pricing.

- Page: https://gpuclouds.fyi/nl/global/online/leafcloud
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Netherlands
- Website: <https://leaf.cloud/>
- Updated: 2026-11-07

### [Nebius](https://gpuclouds.fyi/nl/global/online/nebius.md)

AI cloud offering NVIDIA GPU instances (GB300, B300, GB200, B200, H200, H100, RTX PRO 6000, L40S) on-demand and preemptible, with free managed Kubernetes and Slurm-on-Kubernetes (Soperator).

- Page: https://gpuclouds.fyi/nl/global/online/nebius
- Categories: GPU neoclouds
- Address: Online, Global, Netherlands
- Website: <https://nebius.com/>
- Updated: 2026-10-29

### [Nebul](https://gpuclouds.fyi/nl/global/online/nebul.md)

European private AI cloud offering dedicated NVIDIA GPU infrastructure, clusters, Kubernetes and serverless or dedicated model inference for training and production workloads.

- Page: https://gpuclouds.fyi/nl/global/online/nebul
- Categories: Sovereign and regional GPU, GPU neoclouds, Inference platforms
- Address: Online, Global, Netherlands
- Website: <https://nebul.com/>
- Updated: 2026-10-29

### [Nosana](https://gpuclouds.fyi/nl/global/online/nosana.md)

Open-source distributed GPU cloud on which users rent GPUs from independent hosts on demand for AI inference, training, fine-tuning and rendering, deployed via dashboard or SDK; hosts earn by sharing idle GPUs.

- Page: https://gpuclouds.fyi/nl/global/online/nosana
- Categories: Decentralized GPU, GPU marketplaces
- Address: Online, Global, Netherlands
- Website: <https://nosana.com/>
- Updated: 2026-10-29

### [Tencent Cloud (Cloud GPU Service)](https://gpuclouds.fyi/nl/global/online/tencent-cloud-cloud-gpu-service.md)

Tencent Cloud GPU Service offers GPU computing and rendering instances with NVIDIA T4, V100 and A100 GPUs for deep learning, scientific computing and graphics, available pay-as-you-go or as spot instances.

- Page: https://gpuclouds.fyi/nl/global/online/tencent-cloud-cloud-gpu-service
- Categories: Hyperscalers
- Address: Online, Global, Netherlands
- Website: <https://www.tencentcloud.com/>
- Updated: 2026-10-29

## Online, Global, Panama

### [Aethir](https://gpuclouds.fyi/pa/global/online/aethir.md)

Distributed GPU cloud offering enterprise bare-metal GPU-as-a-Service (Aethir Earth) for AI training and inference, plus low-latency cloud gaming GPUs (Atmosphere), supplied by a network of GPU providers.

- Page: https://gpuclouds.fyi/pa/global/online/aethir
- Categories: Decentralized GPU, Bare-metal GPU clusters
- Address: Online, Global, Panama
- Website: <https://aethir.com/>
- Updated: 2026-10-29

## Online, Global, Russia

### [Cloud.ru (GPU rental)](https://gpuclouds.fyi/ru/global/online/cloud-ru-gpu-rental.md)

Cloud.ru rents dedicated servers, virtual machines and ML services with NVIDIA B300, H100, A100, V100 and A40 GPUs from data centers in Russia, priced in RUB with monthly, hourly or per-minute billing.

- Page: https://gpuclouds.fyi/ru/global/online/cloud-ru-gpu-rental
- Categories: Sovereign and regional GPU, Hyperscalers
- Address: Online, Global, Russia
- Website: <https://cloud.ru/>
- Updated: 2026-11-07

### [Selectel (GPU servers)](https://gpuclouds.fyi/ru/global/online/selectel-gpu-servers.md)

Selectel rents cloud and bare-metal GPU servers with NVIDIA B300, H200, A100, A30, RTX 6000 PRO, L4 and RTX 4090, with pay-as-you-go cloud GPUs, GPU clusters in Managed Kubernetes, and token-based inference.

- Page: https://gpuclouds.fyi/ru/global/online/selectel-gpu-servers
- Categories: Sovereign and regional GPU, Bare-metal GPU clusters
- Address: Online, Global, Russia
- Website: <https://selectel.ru/>
- Updated: 2026-11-07

### [Yandex Cloud (Compute Cloud GPUs)](https://gpuclouds.fyi/ru/global/online/yandex-cloud-compute-cloud-gpus.md)

Yandex Cloud Compute Cloud provides VMs with NVIDIA V100, A100, T4 and T4i GPUs across several GPU platforms, plus GPU clusters for multi-node training, in the Russia region.

- Page: https://gpuclouds.fyi/ru/global/online/yandex-cloud-compute-cloud-gpus
- Categories: Hyperscalers, Sovereign and regional GPU
- Address: Online, Global, Russia
- Website: <https://yandex.cloud/>
- Updated: 2026-10-29

## Online, Global, Saudi Arabia

### [HUMAIN](https://gpuclouds.fyi/sa/global/online/humain.md)

Provides AI cloud infrastructure for model training, fine-tuning and inference, alongside managed HPC services, Kubernetes orchestration and private deployment options.

- Page: https://gpuclouds.fyi/sa/global/online/humain
- Categories: GPU neoclouds, Sovereign and regional GPU
- Address: Online, Global, Saudi Arabia
- Website: <https://www.humain.com/>
- Updated: 2026-10-29

## Online, Global, Sweden

### [evroc](https://gpuclouds.fyi/se/global/online/evroc.md)

European cloud offering GPU virtual machines with one to eight NVIDIA L40S or B200 GPUs for AI training, inference and HPC, with API and CLI access and hourly pay-as-you-go pricing.

- Page: https://gpuclouds.fyi/se/global/online/evroc
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Sweden
- Website: <https://evroc.com/>
- Updated: 2026-11-07

## Online, Global, Singapore

### [Bitdeer AI Cloud](https://gpuclouds.fyi/sg/global/online/bitdeer-ai-cloud.md)

Bitdeer AI Cloud, an NVIDIA Cloud Partner, offers GPU virtual machines, bare metal including GB200 and GB300 NVL72, containers, distributed training jobs and serverless model APIs, with on-demand and 1/3-year pricing.

- Page: https://gpuclouds.fyi/sg/global/online/bitdeer-ai-cloud
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, Singapore
- Website: <https://www.bitdeer.ai/>
- Updated: 2026-10-29

### [BytePlus (GPU Compute Service)](https://gpuclouds.fyi/sg/global/online/byteplus-gpu-compute-service.md)

BytePlus GPU Compute Service provides elastic GPU instances, including HPC clusters with up to 3.2 Tbps RDMA interconnect, for generative AI, autonomous driving, image processing and scientific computing.

- Page: https://gpuclouds.fyi/sg/global/online/byteplus-gpu-compute-service
- Categories: Hyperscalers
- Address: Online, Global, Singapore
- Website: <https://www.byteplus.com/en>
- Updated: 2026-11-07

### [HPC-AI.COM (HPC-AI Tech)](https://gpuclouds.fyi/sg/global/online/hpc-ai-com-hpc-ai-tech.md)

Singapore-based GPU cloud offering on-demand 8-GPU full-machine rentals of NVIDIA B200, H200, H100 and RTX 4090/5090, billed per minute, plus reserved clusters, batch jobs, fine-tuning and model APIs.

- Page: https://gpuclouds.fyi/sg/global/online/hpc-ai-com-hpc-ai-tech
- Categories: GPU neoclouds, Inference platforms
- Address: Online, Global, Singapore
- Website: <https://hpc-ai.com/>
- Updated: 2026-11-07

### [SiliconFlow](https://gpuclouds.fyi/sg/global/online/siliconflow.md)

SiliconFlow is an AI platform offering serverless model inference with per-token pricing for open models such as DeepSeek, Qwen and GLM, plus fine-tuning and reserved dedicated GPUs.

- Page: https://gpuclouds.fyi/sg/global/online/siliconflow
- Categories: Inference platforms
- Address: Online, Global, Singapore
- Website: <https://www.siliconflow.com/>
- Updated: 2026-11-07

### [Singtel RE:AI](https://gpuclouds.fyi/sg/global/online/singtel-re-ai.md)

RE:AI is the sovereign AI cloud business of Singtel Digital InfraCo, offering GPU-as-a-Service and AI-as-a-Service on Nxera data centres, including NVIDIA DGX GB300 capacity.

- Page: https://gpuclouds.fyi/sg/global/online/singtel-re-ai
- Categories: Sovereign and regional GPU
- Address: Online, Global, Singapore
- Website: <https://www.singtel.com/digital-infraco>
- Updated: 2026-11-07

### [Spheron](https://gpuclouds.fyi/sg/global/online/spheron.md)

GPU cloud marketplace aggregating capacity from multiple data center providers into one dashboard, offering on-demand, spot and reserved GPU instances plus managed LLM inference endpoints.

- Page: https://gpuclouds.fyi/sg/global/online/spheron
- Categories: GPU marketplaces, Inference platforms
- Address: Online, Global, Singapore
- Website: <https://www.spheron.network/>
- Updated: 2026-11-07

## Online, Global, Taiwan

### [Glows.ai](https://gpuclouds.fyi/tw/global/online/glows-ai.md)

Globally distributed GPU cloud for AI training and inference offering on-demand GPU instances, bare-metal and virtualized clusters, inference endpoints and data storage, with per-second billing.

- Page: https://gpuclouds.fyi/tw/global/online/glows-ai
- Categories: GPU neoclouds, Inference platforms
- Address: Online, Global, Taiwan
- Website: <https://glows.ai/>
- Updated: 2026-11-07

### [Taiwan AI Cloud](https://gpuclouds.fyi/tw/global/online/taiwan-ai-cloud.md)

Offers NVIDIA H100 cloud GPU instances and dedicated GPU clusters for AI training and inference, with hourly billing and data centers in Taiwan.

- Page: https://gpuclouds.fyi/tw/global/online/taiwan-ai-cloud
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Taiwan
- Website: <https://www.twcloud.ai/>
- Updated: 2026-11-07

## Online, Global, United States

### [Anyscale](https://gpuclouds.fyi/us/global/online/anyscale.md)

Managed Ray platform for distributed training, batch inference and model serving, offering Anyscale-hosted GPU compute (T4, L4, A10G, A100, H/B/GB families) or deployment in the customer's own cloud.

- Page: https://gpuclouds.fyi/us/global/online/anyscale
- Categories: Inference platforms, Serverless GPU
- Address: Online, Global, United States
- Website: <https://www.anyscale.com/>
- Updated: 2026-11-07

### [Atlantic.Net](https://gpuclouds.fyi/us/global/online/atlantic-net.md)

Hosting provider and NVIDIA cloud partner offering on-demand cloud GPU and dedicated GPU servers with NVIDIA H100 NVL, L40S and A2 GPUs, including HIPAA-compliant GPU hosting and an LLM inference cloud.

- Page: https://gpuclouds.fyi/us/global/online/atlantic-net
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://www.atlantic.net/>
- Updated: 2026-11-07

### [Atlas Cloud](https://gpuclouds.fyi/us/global/online/atlas-cloud.md)

Atlas Cloud offers a single OpenAI-compatible API for 400+ image, video, audio and language models, plus on-demand GPU instances with NVIDIA H100 SXM and H200 billed per GPU-hour.

- Page: https://gpuclouds.fyi/us/global/online/atlas-cloud
- Categories: Inference platforms, GPU neoclouds
- Address: Online, Global, United States
- Website: <https://www.atlascloud.ai/>
- Updated: 2026-11-07

### [AWS (Amazon EC2)](https://gpuclouds.fyi/us/global/online/aws-amazon-ec2.md)

Amazon Web Services offers EC2 accelerated computing instances with NVIDIA GPUs, including P5 (H100), P5e/P5en (H200), P6 (B200/B300) and P4d (A100), billed per second on-demand, with Spot and Capacity Blocks.

- Page: https://gpuclouds.fyi/us/global/online/aws-amazon-ec2
- Categories: Hyperscalers
- Address: Online, Global, United States
- Website: <https://aws.amazon.com/>
- Updated: 2026-10-29

### [Baseten](https://gpuclouds.fyi/us/global/online/baseten.md)

AI inference platform offering model APIs, dedicated model deployments and training on GPU instances from T4 to B200, billed by the minute, with options to deploy in Baseten cloud, customer VPC or hybrid.

- Page: https://gpuclouds.fyi/us/global/online/baseten
- Categories: Inference platforms, Serverless GPU
- Address: Online, Global, United States
- Website: <https://www.baseten.co/>
- Updated: 2026-10-29

### [Beam](https://gpuclouds.fyi/us/global/online/beam.md)

Cloud platform offering serverless GPU endpoints billed by the millisecond, persistent VMs, sandboxes, on-demand bare-metal or VM GPU machines, reserved InfiniBand clusters, and bring-your-own-cloud.

- Page: https://gpuclouds.fyi/us/global/online/beam
- Categories: Serverless GPU, GPU neoclouds
- Address: Online, Global, United States
- Website: <https://www.beam.cloud/>
- Updated: 2026-10-29

### [Cerebras](https://gpuclouds.fyi/us/global/online/cerebras.md)

Inference cloud serving language models through public API endpoints and dedicated deployments on Cerebras wafer-scale hardware, with training and fine-tuning options.

- Page: https://gpuclouds.fyi/us/global/online/cerebras
- Categories: Inference platforms
- Address: Online, Global, United States
- Website: <https://www.cerebras.ai/>
- Updated: 2026-11-07

### [Cerebrium](https://gpuclouds.fyi/us/global/online/cerebrium.md)

Serverless AI infrastructure platform for deploying LLM, voice and image workloads on GPUs from T4 to B200, scaling with demand and charging for compute time measured in seconds.

- Page: https://gpuclouds.fyi/us/global/online/cerebrium
- Categories: Serverless GPU, Inference platforms
- Address: Online, Global, United States
- Website: <https://www.cerebrium.ai/>
- Updated: 2026-11-07

### [ChronoScale](https://gpuclouds.fyi/us/global/online/chronoscale.md)

AI infrastructure company spun out of Applied Digital's cloud business, providing dedicated GPU environments for training, inference and HPC, contracted by the megawatt across US, APAC and Nordic campuses.

- Page: https://gpuclouds.fyi/us/global/online/chronoscale
- Categories: Bare-metal GPU clusters, GPU neoclouds
- Address: Online, Global, United States
- Website: <https://www.chronoscale.com/>
- Updated: 2026-10-29

### [Cloudflare Workers AI](https://gpuclouds.fyi/us/global/online/cloudflare-workers-ai.md)

Serverless inference service running 50+ open-source models on GPUs across Cloudflare's global network, priced per Neuron ($0.011 per 1,000) with a free daily allocation of 10,000 Neurons.

- Page: https://gpuclouds.fyi/us/global/online/cloudflare-workers-ai
- Categories: Inference platforms, Serverless GPU
- Address: Online, Global, United States
- Website: <https://www.cloudflare.com/>
- Updated: 2026-10-29

### [CloudRift](https://gpuclouds.fyi/us/global/online/cloudrift.md)

GPU rental platform and sovereign AI infrastructure software offering on-demand and reserved NVIDIA and AMD GPU instances (VMs, containers, bare metal) billed per second, plus a control plane for GPU operators.

- Page: https://gpuclouds.fyi/us/global/online/cloudrift
- Categories: GPU neoclouds, Sovereign and regional GPU
- Address: Online, Global, United States
- Website: <https://www.cloudrift.ai/>
- Updated: 2026-11-07

### [CoreWeave](https://gpuclouds.fyi/us/global/online/coreweave.md)

AI cloud offering NVIDIA GPU instances (GB300/GB200 NVL72, B200, H200, H100, L40S, A100 and more) on demand and spot, plus managed Kubernetes (CKS), SUNK, storage and inference services.

- Page: https://gpuclouds.fyi/us/global/online/coreweave
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://www.coreweave.com/>
- Updated: 2026-11-07

### [Crusoe](https://gpuclouds.fyi/us/global/online/crusoe.md)

AI cloud offering NVIDIA and AMD GPU instances (GB300, B300, H200, H100, A100, L40S, MI300X) on spot, on-demand or reserved terms, plus managed Kubernetes, Slurm, and serverless inference and fine-tuning.

- Page: https://gpuclouds.fyi/us/global/online/crusoe
- Categories: GPU neoclouds, Inference platforms
- Address: Online, Global, United States
- Website: <https://www.crusoe.ai/>
- Updated: 2026-10-29

### [DeepInfra](https://gpuclouds.fyi/us/global/online/deepinfra.md)

Inference platform serving open and proprietary models via API priced per token or per unit, plus custom model deployments on dedicated A100, H100, H200, B200 and B300 GPUs billed by the minute.

- Page: https://gpuclouds.fyi/us/global/online/deepinfra
- Categories: Inference platforms, Serverless GPU
- Address: Online, Global, United States
- Website: <https://deepinfra.com/>
- Updated: 2026-11-07

### [DigitalOcean GPU Droplets](https://gpuclouds.fyi/us/global/online/digitalocean-gpu-droplets.md)

Cloud provider offering GPU Droplets with NVIDIA (B300, H200, H100, L40S, RTX 6000/4000 Ada) and AMD Instinct (MI355X, MI350X, MI325X, MI300X) GPUs on-demand, spot or 12-month reserved, billed per second.

- Page: https://gpuclouds.fyi/us/global/online/digitalocean-gpu-droplets
- Categories: GPU neoclouds
- Address: Online, Global, United States
- Website: <https://www.digitalocean.com/>
- Updated: 2026-10-29

### [fal](https://gpuclouds.fyi/us/global/online/fal.md)

Generative media platform offering model APIs for image, video, audio and 3D generation billed per output, plus serverless compute for deploying custom apps on its GPU fleet.

- Page: https://gpuclouds.fyi/us/global/online/fal
- Categories: Inference platforms, Serverless GPU
- Address: Online, Global, United States
- Website: <https://fal.ai/>
- Updated: 2026-11-07

### [Featherless](https://gpuclouds.fyi/us/global/online/featherless.md)

Serverless inference platform for open-weight models with an OpenAI-compatible API, monthly chat subscriptions, metered developer plans and dedicated GPU deployments.

- Page: https://gpuclouds.fyi/us/global/online/featherless
- Categories: Inference platforms, Serverless GPU
- Address: Online, Global, United States
- Website: <https://featherless.ai/>
- Updated: 2026-11-07

### [Fireworks AI](https://gpuclouds.fyi/us/global/online/fireworks-ai.md)

Generative AI inference platform offering serverless per-token model APIs, fine-tuning and training, and on-demand GPU deployments on H100 to GB300 GPUs priced per GPU second.

- Page: https://gpuclouds.fyi/us/global/online/fireworks-ai
- Categories: Inference platforms
- Address: Online, Global, United States
- Website: <https://fireworks.ai/>
- Updated: 2026-10-29

### [Fluence](https://gpuclouds.fyi/us/global/online/fluence.md)

Fluence Cloud lets teams deploy on-demand GPU and CPU instances sourced from multiple providers via console or API, with zero egress fees and custom GPU cluster requests; backed by the Fluence protocol.

- Page: https://gpuclouds.fyi/us/global/online/fluence
- Categories: Decentralized GPU, GPU marketplaces
- Address: Online, Global, United States
- Website: <https://fluence.ai/>
- Updated: 2026-10-29

### [FriendliAI](https://gpuclouds.fyi/us/global/online/friendliai.md)

AI inference platform offering model APIs, on-demand dedicated GPU endpoints on A100 to B300 GPUs billed per second, and bring-your-own-GPU deployments, with VPC and on-prem options for enterprises.

- Page: https://gpuclouds.fyi/us/global/online/friendliai
- Categories: Inference platforms
- Address: Online, Global, United States
- Website: <https://friendli.ai/>
- Updated: 2026-11-07

### [GMI Cloud](https://gpuclouds.fyi/us/global/online/gmi-cloud.md)

GMI Cloud provides dedicated single-tenant NVIDIA GPUs (H100, H200, B200, GB200, GB300) in GMI-operated data centers, on-demand or reserved, priced per GPU-hour, plus a model inference platform.

- Page: https://gpuclouds.fyi/us/global/online/gmi-cloud
- Categories: GPU neoclouds, Inference platforms
- Address: Online, Global, United States
- Website: <https://www.gmicloud.ai/>
- Updated: 2026-10-29

### [Google Cloud (Compute Engine GPUs)](https://gpuclouds.fyi/us/global/online/google-cloud-compute-engine-gpus.md)

Google Cloud Compute Engine offers accelerator-optimized VMs (A4X, A4, A3, A2, G4, G2) with NVIDIA GPUs from GB300 and B200 to H100, A100 and L4, available on-demand, as Spot VMs, or with committed use discounts.

- Page: https://gpuclouds.fyi/us/global/online/google-cloud-compute-engine-gpus
- Categories: Hyperscalers
- Address: Online, Global, United States
- Website: <https://cloud.google.com/>
- Updated: 2026-10-29

### [GPU Mart](https://gpuclouds.fyi/us/global/online/gpu-mart.md)

GPU hosting brand of Database Mart LLC offering dedicated GPU servers and GPU VPS (H100, A100, RTX Pro 6000, RTX 5090/4090 and others) from owned data centers in Dallas, Texas and Kansas City, Missouri.

- Page: https://gpuclouds.fyi/us/global/online/gpu-mart
- Categories: Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://www.gpu-mart.com/>
- Updated: 2026-10-29

### [Groq](https://gpuclouds.fyi/us/global/online/groq.md)

Inference cloud using Groq LPU accelerators, with dedicated bare-metal infrastructure, managed inference and enterprise controls, plus an OpenAI-compatible API.

- Page: https://gpuclouds.fyi/us/global/online/groq
- Categories: Inference platforms
- Address: Online, Global, United States
- Website: <https://groq.com/>
- Updated: 2026-10-29

### [Highrise AI (Hut 8)](https://gpuclouds.fyi/us/global/online/highrise-ai-hut-8.md)

Hut 8 subsidiary providing GPU pods and dedicated clusters for AI training and inference, with bare-metal access, managed Kubernetes and reserved capacity.

- Page: https://gpuclouds.fyi/us/global/online/highrise-ai-hut-8
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://highrise.ai/>
- Updated: 2026-10-29

### [Hivelocity](https://gpuclouds.fyi/us/global/online/hivelocity.md)

Tampa-based bare-metal and dedicated server provider whose instant dedicated server lineup includes configurations with an NVIDIA L4 24GB GPU, with custom GPU server builds available via sales.

- Page: https://gpuclouds.fyi/us/global/online/hivelocity
- Categories: Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://www.hivelocity.net/>
- Updated: 2026-10-29

### [Hugging Face Inference Endpoints](https://gpuclouds.fyi/us/global/online/hugging-face-inference-endpoints.md)

Managed service from Hugging Face for deploying models on dedicated, autoscaling endpoints on AWS, GCP or Azure infrastructure, with GPU instances (T4, L4, A10G, L40S, A100, H100, H200) charged by the minute.

- Page: https://gpuclouds.fyi/us/global/online/hugging-face-inference-endpoints
- Categories: Inference platforms
- Address: Online, Global, United States
- Website: <https://huggingface.co/>
- Updated: 2026-10-29

### [Hydra Host](https://gpuclouds.fyi/us/global/online/hydra-host.md)

Bare-metal GPU platform (Brokkr) aggregating capacity from independent data center operators across 40+ data centers, offering NVIDIA and AMD GPUs on on-demand, interruptible or reserved multi-year terms.

- Page: https://gpuclouds.fyi/us/global/online/hydra-host
- Categories: Bare-metal GPU clusters, GPU marketplaces
- Address: Online, Global, United States
- Website: <https://hydrahost.com/>
- Updated: 2026-11-07

### [Hyperbolic](https://gpuclouds.fyi/us/global/online/hyperbolic.md)

AI compute platform offering on-demand and reserved NVIDIA H100, H200 and B200 GPUs sourced from a supplier network, managed Kubernetes with GPU worker nodes, and an OpenAI-compatible inference API.

- Page: https://gpuclouds.fyi/us/global/online/hyperbolic
- Categories: GPU marketplaces, Inference platforms
- Address: Online, Global, United States
- Website: <https://www.hyperbolic.ai/>
- Updated: 2026-11-07

### [IBM Cloud GPUs](https://gpuclouds.fyi/us/global/online/ibm-cloud-gpus.md)

IBM Cloud provides on-demand GPU and AI accelerator instances, including NVIDIA B300, H200, H100, A100, L40S and L4, AMD Instinct MI300X and Intel Gaudi 3, via Virtual Servers for VPC, Red Hat OpenShift and Serverless.

- Page: https://gpuclouds.fyi/us/global/online/ibm-cloud-gpus
- Categories: Hyperscalers
- Address: Online, Global, United States
- Website: <https://www.ibm.com/cloud>
- Updated: 2026-11-07

### [Inference.net](https://gpuclouds.fyi/us/global/online/inference-net.md)

Inference infrastructure for open-source, custom and fine-tuned models, offering dedicated model deployments on B200 GPU instances priced per hour, plus observability, tracing, evaluation and custom model training.

- Page: https://gpuclouds.fyi/us/global/online/inference-net
- Categories: Inference platforms
- Address: Online, Global, United States
- Website: <https://inference.net/>
- Updated: 2026-10-29

### [io.net](https://gpuclouds.fyi/us/global/online/io-net.md)

Decentralized GPU cloud aggregating GPUs from many suppliers; offers on-demand GPU clusters (Ray, Kubernetes, containers, bare metal) plus io.intelligence model inference APIs.

- Page: https://gpuclouds.fyi/us/global/online/io-net
- Categories: Decentralized GPU, GPU marketplaces, Inference platforms
- Address: Online, Global, United States
- Website: <https://io.net/>
- Updated: 2026-10-29

### [ionstream](https://gpuclouds.fyi/us/global/online/ionstream.md)

GPU-as-a-service provider offering single-tenant bare-metal NVIDIA HGX B200 and B300 servers, provisioned and networked through its Gravity platform, with reserved GPU capacity for training and inference.

- Page: https://gpuclouds.fyi/us/global/online/ionstream
- Categories: Bare-metal GPU clusters, GPU neoclouds
- Address: Online, Global, United States
- Website: <https://ionstream.ai/>
- Updated: 2026-11-07

### [Lambda](https://gpuclouds.fyi/us/global/online/lambda.md)

GPU cloud offering on-demand NVIDIA B200, H100, A100, GH200 and other GPU instances, 1-Click Clusters of 16 to 2,000+ GPUs, private cloud, and managed Kubernetes and Slurm.

- Page: https://gpuclouds.fyi/us/global/online/lambda
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://lambda.ai/>
- Updated: 2026-10-29

### [Latitude.sh](https://gpuclouds.fyi/us/global/online/latitude-sh.md)

Bare-metal cloud platform offering Metal GPU instances and dedicated GPU clusters (NVIDIA HGX B300, RTX PRO 6000, H100) with hourly, monthly or yearly billing, plus a managed Kubernetes service.

- Page: https://gpuclouds.fyi/us/global/online/latitude-sh
- Categories: Bare-metal GPU clusters, GPU neoclouds
- Address: Online, Global, United States
- Website: <https://www.latitude.sh/>
- Updated: 2026-11-07

### [Lightning AI](https://gpuclouds.fyi/us/global/online/lightning-ai.md)

AI development platform (Lightning Studios) providing cloud GPU and CPU machines, from T4 and L4 to A100, H100, H200 and B200, billed by the second with interruptible pricing and free starter GPU hours.

- Page: https://gpuclouds.fyi/us/global/online/lightning-ai
- Categories: GPU neoclouds, Serverless GPU
- Address: Online, Global, United States
- Website: <https://lightning.ai/>
- Updated: 2026-11-07

### [Lilypad Network](https://gpuclouds.fyi/us/global/online/lilypad-network.md)

Decentralized compute network and AI services platform where developers run AI inference jobs via an inference API or CLI on a global network of GPU resource providers, and publish and monetize models.

- Page: https://gpuclouds.fyi/us/global/online/lilypad-network
- Categories: Decentralized GPU, Inference platforms
- Address: Online, Global, United States
- Website: <https://lilypad.tech/>
- Updated: 2026-11-07

### [Liquid Web](https://gpuclouds.fyi/us/global/online/liquid-web.md)

Managed hosting company offering bare-metal GPU servers with NVIDIA L4, L40S, H100 and H200 GPUs in single and dual-GPU configurations, provisioned on demand for ML/DL training and inference.

- Page: https://gpuclouds.fyi/us/global/online/liquid-web
- Categories: Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://www.liquidweb.com/>
- Updated: 2026-10-29

### [Massed Compute](https://gpuclouds.fyi/us/global/online/massed-compute.md)

GPU cloud that owns and operates its infrastructure, renting NVIDIA GPU instances (B300, B200, H200, H100, A100, L40S, RTX PRO 6000 and more) by the hour with no contracts, plus bare metal and an API.

- Page: https://gpuclouds.fyi/us/global/online/massed-compute
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://massedcompute.com/>
- Updated: 2026-10-29

### [Microsoft Azure (GPU VMs)](https://gpuclouds.fyi/us/global/online/microsoft-azure-gpu-vms.md)

Microsoft Azure offers GPU-optimized virtual machine families (NC, ND, NV) with NVIDIA GB300, GB200, H200, H100, A100, RTX PRO 6000, A10 and T4 GPUs and AMD MI300X, available pay-as-you-go, as Spot VMs or reserved.

- Page: https://gpuclouds.fyi/us/global/online/microsoft-azure-gpu-vms
- Categories: Hyperscalers
- Address: Online, Global, United States
- Website: <https://azure.microsoft.com/>
- Updated: 2026-10-29

### [Mithril (formerly Foundry / ML Foundry)](https://gpuclouds.fyi/us/global/online/mithril-formerly-foundry-ml-foundry.md)

Self-serve GPU cloud from Foundry Technologies offering NVIDIA H100 and H200 instances via market-priced spot bids and flexible reservations, with API, CLI and Kubernetes cluster support.

- Page: https://gpuclouds.fyi/us/global/online/mithril-formerly-foundry-ml-foundry
- Categories: GPU neoclouds, GPU marketplaces
- Address: Online, Global, United States
- Website: <https://mithril.ai/>
- Updated: 2026-10-29

### [Modal](https://gpuclouds.fyi/us/global/online/modal.md)

Serverless AI infrastructure platform for running inference, training, batch jobs and sandboxes on GPUs and CPUs that spin up in seconds, billed per second for actual compute time.

- Page: https://gpuclouds.fyi/us/global/online/modal
- Categories: Serverless GPU, Inference platforms
- Address: Online, Global, United States
- Website: <https://modal.com/>
- Updated: 2026-10-29

### [Novita AI](https://gpuclouds.fyi/us/global/online/novita-ai.md)

AI cloud offering model inference APIs for LLM, image, video and audio, plus GPU instances and serverless GPUs across 14 global regions, with on-demand and spot options and per-second billing.

- Page: https://gpuclouds.fyi/us/global/online/novita-ai
- Categories: Inference platforms, GPU neoclouds, Serverless GPU
- Address: Online, Global, United States
- Website: <https://novita.ai/>
- Updated: 2026-11-07

### [NVIDIA Brev](https://gpuclouds.fyi/us/global/online/nvidia-brev.md)

NVIDIA platform giving access to GPU virtual machines preconfigured with CUDA, Python, Docker and JupyterLab across multiple cloud providers, plus shareable Launchables bundling hardware, software and code.

- Page: https://gpuclouds.fyi/us/global/online/nvidia-brev
- Categories: GPU marketplaces
- Address: Online, Global, United States
- Website: <https://brev.nvidia.com/>
- Updated: 2026-10-29

### [NVIDIA DGX Cloud Lepton](https://gpuclouds.fyi/us/global/online/nvidia-dgx-cloud-lepton.md)

NVIDIA platform connecting developers to GPU compute across cloud partners and marketplaces, with a unified workflow for AI development, training and inference across providers.

- Page: https://gpuclouds.fyi/us/global/online/nvidia-dgx-cloud-lepton
- Categories: GPU marketplaces, Inference platforms
- Address: Online, Global, United States
- Website: <https://lepton.ai/>
- Updated: 2026-10-29

### [Oblivus](https://gpuclouds.fyi/us/global/online/oblivus.md)

Offers on-demand and reserved NVIDIA GPU instances for AI training, inference and rendering, with per-minute billing, an API and deployment in Montreal and Oslo.

- Page: https://gpuclouds.fyi/us/global/online/oblivus
- Categories: GPU neoclouds
- Address: Online, Global, United States
- Website: <https://oblivus.com/>
- Updated: 2026-11-07

### [Oracle Cloud Infrastructure (OCI) GPU](https://gpuclouds.fyi/us/global/online/oracle-cloud-infrastructure-oci-gpu.md)

Oracle Cloud Infrastructure offers bare metal and VM GPU instances with NVIDIA GB300, GB200, B300, B200, H200, H100, A100, L40S and A10 GPUs and AMD MI355X/MI300X, priced per GPU-hour, scalable to OCI Supercluster.

- Page: https://gpuclouds.fyi/us/global/online/oracle-cloud-infrastructure-oci-gpu
- Categories: Hyperscalers, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://www.oracle.com/cloud/>
- Updated: 2026-10-29

### [Paperspace (DigitalOcean)](https://gpuclouds.fyi/us/global/online/paperspace-digitalocean.md)

GPU cloud now part of DigitalOcean, offering on-demand NVIDIA GPU machines (H100, A100, A6000, A5000, A4000, V100 and others) billed hourly, plus Gradient notebooks with a free GPU tier.

- Page: https://gpuclouds.fyi/us/global/online/paperspace-digitalocean
- Categories: GPU neoclouds
- Address: Online, Global, United States
- Website: <https://www.paperspace.com/>
- Updated: 2026-11-07

### [Parasail](https://gpuclouds.fyi/us/global/online/parasail.md)

AI inference platform offering per-token serverless endpoints for open models, self-service batch processing, and dedicated deployments on reserved NVIDIA GPUs priced per GPU-hour and billed by the minute.

- Page: https://gpuclouds.fyi/us/global/online/parasail
- Categories: Inference platforms
- Address: Online, Global, United States
- Website: <https://www.parasail.io/>
- Updated: 2026-11-07

### [Phala Cloud (GPU TEE)](https://gpuclouds.fyi/us/global/online/phala-cloud-gpu-tee.md)

Confidential AI cloud running containers, private LLM inference and GPU jobs inside hardware TEEs (Intel TDX, NVIDIA confidential computing) with attestation; offers H100, H200 and B300 GPU TEE capacity.

- Page: https://gpuclouds.fyi/us/global/online/phala-cloud-gpu-tee
- Categories: GPU neoclouds, Decentralized GPU, Inference platforms
- Address: Online, Global, United States
- Website: <https://phala.com/>
- Updated: 2026-10-29

### [phoenixNAP](https://gpuclouds.fyi/us/global/online/phoenixnap.md)

Infrastructure provider whose Bare Metal Cloud offers on-demand dedicated GPU servers with dual Intel Max 1100 GPUs, provisioned via API, CLI or web portal, with hourly, monthly and reserved billing.

- Page: https://gpuclouds.fyi/us/global/online/phoenixnap
- Categories: Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://phoenixnap.com/>
- Updated: 2026-11-07

### [Prime Intellect](https://gpuclouds.fyi/us/global/online/prime-intellect.md)

AI training, inference and compute platform offering on-demand access to 1-256 GPUs across clouds, reserved clusters quoted from 50+ providers, a spot resale market, plus hosted RL training and a model API.

- Page: https://gpuclouds.fyi/us/global/online/prime-intellect
- Categories: GPU marketplaces, Inference platforms
- Address: Online, Global, United States
- Website: <https://www.primeintellect.ai/>
- Updated: 2026-10-29

### [QuickPod](https://gpuclouds.fyi/us/global/online/quickpod.md)

Distributed GPU and CPU marketplace for renting compute from host machines, with templates such as Jupyter for PyTorch and TensorFlow, persistent storage, clusters and serverless endpoints.

- Page: https://gpuclouds.fyi/us/global/online/quickpod
- Categories: GPU marketplaces, Decentralized GPU
- Address: Online, Global, United States
- Website: <https://quickpod.io/>
- Updated: 2026-10-29

### [QumulusAI](https://gpuclouds.fyi/us/global/online/qumulusai.md)

Atlanta-based NVIDIA Cloud Partner that designs and deploys single-tenant GPU environments with committed capacity, offering NVIDIA B300, B200, H200, H100, RTX Pro 6000 and L4 GPUs from dedicated nodes to full clusters.

- Page: https://gpuclouds.fyi/us/global/online/qumulusai
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://www.qumulusai.com/>
- Updated: 2026-11-07

### [Rackspace Spot](https://gpuclouds.fyi/us/global/online/rackspace-spot.md)

Rackspace's auction-based cloud where users bid per node-hour for VMs and managed Kubernetes capacity, including GPU server classes with NVIDIA H100 and A30 GPUs; on-demand rates are also listed.

- Page: https://gpuclouds.fyi/us/global/online/rackspace-spot
- Categories: GPU marketplaces, GPU neoclouds
- Address: Online, Global, United States
- Website: <https://spot.rackspace.com/>
- Updated: 2026-11-07

### [Render Network](https://gpuclouds.fyi/us/global/online/render-network.md)

Distributed GPU rendering network built on OTOY software that runs a compute marketplace between GPU providers and requestors, supporting Octane, Redshift and Blender Cycles rendering and generative AI tools.

- Page: https://gpuclouds.fyi/us/global/online/render-network
- Categories: Decentralized GPU
- Address: Online, Global, United States
- Website: <https://rendernetwork.com/>
- Updated: 2026-10-29

### [Rescale](https://gpuclouds.fyi/us/global/online/rescale.md)

Cloud HPC and AI platform for engineering simulation that provides access to CPU and NVIDIA GPU architectures (including A100 and V100) deployed across global cloud providers through a unified platform.

- Page: https://gpuclouds.fyi/us/global/online/rescale
- Categories: GPU neoclouds
- Address: Online, Global, United States
- Website: <https://rescale.com/>
- Updated: 2026-10-29

### [Runpod](https://gpuclouds.fyi/us/global/online/runpod.md)

GPU cloud offering on-demand GPU pods across 30+ regions, serverless GPU endpoints for API inference, and multi-node GPU clusters, with Community Cloud and Secure Cloud tiers.

- Page: https://gpuclouds.fyi/us/global/online/runpod
- Categories: GPU neoclouds, GPU marketplaces, Serverless GPU
- Address: Online, Global, United States
- Website: <https://www.runpod.io/>
- Updated: 2026-11-07

### [SaladCloud](https://gpuclouds.fyi/us/global/online/saladcloud.md)

Distributed GPU cloud running containers on idle consumer RTX-class GPUs in homes and small businesses worldwide, with four priority tiers and per-second billing; also offers reserved RTX PRO 6000 servers.

- Page: https://gpuclouds.fyi/us/global/online/saladcloud
- Categories: Decentralized GPU, Serverless GPU
- Address: Online, Global, United States
- Website: <https://salad.com/>
- Updated: 2026-10-29

### [SambaNova](https://gpuclouds.fyi/us/global/online/sambanova.md)

SambaCloud provides language model inference APIs priced per token on SambaNova RDU hardware; SambaStack supports deployments using the same inference technology.

- Page: https://gpuclouds.fyi/us/global/online/sambanova
- Categories: Inference platforms
- Address: Online, Global, United States
- Website: <https://sambanova.ai/>
- Updated: 2026-10-29

### [Saturn Cloud](https://gpuclouds.fyi/us/global/online/saturn-cloud.md)

Hosted data science and ML platform providing Jupyter, IDE and job resources on cloud CPUs and GPUs (H100 via Nebius; T4, V100, A10G via AWS), charged per hour of use with a free plan to start.

- Page: https://gpuclouds.fyi/us/global/online/saturn-cloud
- Categories: GPU neoclouds
- Address: Online, Global, United States
- Website: <https://saturncloud.io/>
- Updated: 2026-11-07

### [Segmind](https://gpuclouds.fyi/us/global/online/segmind.md)

Generative AI platform offering serverless pay-as-you-go APIs for image, video and other media models, Pixelflow workflows, fine-tuning, and dedicated GPU endpoints priced per hour or per second.

- Page: https://gpuclouds.fyi/us/global/online/segmind
- Categories: Inference platforms
- Address: Online, Global, United States
- Website: <https://www.segmind.com/>
- Updated: 2026-11-07

### [SF Compute](https://gpuclouds.fyi/us/global/online/sf-compute.md)

GPU cloud that builds data centers and GPU clusters, offering VMs, bare metal and managed Slurm, with a compute market that lets customers resell reserved capacity they do not use.

- Page: https://gpuclouds.fyi/us/global/online/sf-compute
- Categories: GPU neoclouds, Bare-metal GPU clusters, GPU marketplaces
- Address: Online, Global, United States
- Website: <https://sfcompute.com/>
- Updated: 2026-10-29

### [Shadeform](https://gpuclouds.fyi/us/global/online/shadeform.md)

GPU cloud marketplace for provisioning GPU instances across 30+ cloud providers and 100+ regions from one console and API with centralized billing; reserved contracts also available.

- Page: https://gpuclouds.fyi/us/global/online/shadeform
- Categories: GPU marketplaces
- Address: Online, Global, United States
- Website: <https://shadeform.com/>
- Updated: 2026-10-29

### [Super Protocol](https://gpuclouds.fyi/us/global/online/super-protocol.md)

Confidential AI platform: Super Swarm runs AI workloads in CPU/GPU TEEs across clouds and data centers, and Confidential Router serves open models via an OpenAI-compatible API on confidential GPUs.

- Page: https://gpuclouds.fyi/us/global/online/super-protocol
- Categories: Inference platforms, Decentralized GPU
- Address: Online, Global, United States
- Website: <https://superprotocol.com/>
- Updated: 2026-11-07

### [Targon](https://gpuclouds.fyi/us/global/online/targon.md)

On-demand NVIDIA GPU cloud for training and inference, offering virtual machines with up to eight GPUs, Intel TDX confidential VMs, dedicated bare-metal servers and reserved capacity.

- Page: https://gpuclouds.fyi/us/global/online/targon
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://targon.com/>
- Updated: 2026-10-29

### [TensorDock](https://gpuclouds.fyi/us/global/online/tensordock.md)

Marketplace GPU cloud of vetted independent hosts offering KVM virtual machines with dedicated GPUs and root access across 100+ locations in over 20 countries, billed pay-as-you-go.

- Page: https://gpuclouds.fyi/us/global/online/tensordock
- Categories: GPU marketplaces
- Address: Online, Global, United States
- Website: <https://www.tensordock.com/>
- Updated: 2026-11-07

### [TensorWave](https://gpuclouds.fyi/us/global/online/tensorwave.md)

AI cloud built on AMD Instinct accelerators (MI455X, MI355X, MI325X, MI300X) offering bare-metal GPU clusters, managed Kubernetes and Slurm, network storage, and inference support.

- Page: https://gpuclouds.fyi/us/global/online/tensorwave
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://tensorwave.com/>
- Updated: 2026-10-29

### [Theta EdgeCloud](https://gpuclouds.fyi/us/global/online/theta-edgecloud.md)

Hybrid cloud-edge platform from Theta offering on-demand GPU compute, hosted AI model inference, AI agents and video/rendering developer tools; contributors can supply spare GPU capacity for rewards.

- Page: https://gpuclouds.fyi/us/global/online/theta-edgecloud
- Categories: Decentralized GPU, Inference platforms
- Address: Online, Global, United States
- Website: <https://www.thetaedgecloud.com/>
- Updated: 2026-11-07

### [Thunder Compute](https://gpuclouds.fyi/us/global/online/thunder-compute.md)

GPU cloud offering persistent on-demand GPU instances (RTX A6000, L40, A100 80GB, H100 SXM) billed per minute, and short-lived GPU sandboxes for agents billed per second.

- Page: https://gpuclouds.fyi/us/global/online/thunder-compute
- Categories: GPU neoclouds, Serverless GPU
- Address: Online, Global, United States
- Website: <https://www.thundercompute.com/>
- Updated: 2026-11-07

### [Together AI](https://gpuclouds.fyi/us/global/online/together-ai.md)

AI platform offering serverless and dedicated inference APIs for open models, fine-tuning, and self-serve GPU clusters (H100, H200, B200, B300, GB200/GB300) with managed Kubernetes or Slurm.

- Page: https://gpuclouds.fyi/us/global/online/together-ai
- Categories: Inference platforms, GPU neoclouds
- Address: Online, Global, United States
- Website: <https://www.together.ai/>
- Updated: 2026-10-29

### [VALDI](https://gpuclouds.fyi/us/global/online/valdi.md)

Provides on-demand GPU virtual machines for AI training, inference and other compute workloads, with per-second usage billing and Ubuntu or Windows environments.

- Page: https://gpuclouds.fyi/us/global/online/valdi
- Categories: GPU neoclouds
- Address: Online, Global, United States
- Website: <https://valdi.ai/>
- Updated: 2026-11-07

### [Vast.ai](https://gpuclouds.fyi/us/global/online/vast-ai.md)

GPU marketplace where prices are set by supply and demand across 40+ data centers, offering on-demand, interruptible and reserved instances, plus serverless endpoints and multi-node clusters.

- Page: https://gpuclouds.fyi/us/global/online/vast-ai
- Categories: GPU marketplaces, Serverless GPU
- Address: Online, Global, United States
- Website: <https://vast.ai/>
- Updated: 2026-10-29

### [Voltage Park](https://gpuclouds.fyi/us/global/online/voltage-park.md)

GPU cloud (now merged with Lightning AI) offering self-serve on-demand NVIDIA HGX H100 GPUs, dedicated reserved H200/Blackwell clusters, bare-metal access and managed Kubernetes.

- Page: https://gpuclouds.fyi/us/global/online/voltage-park
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://www.voltagepark.com/>
- Updated: 2026-11-07

### [WhiteFiber](https://gpuclouds.fyi/us/global/online/whitefiber.md)

AI and HPC infrastructure provider offering NVIDIA B300, GB300 and VR200 GPU clusters as reserved GPU cloud capacity, turn-key private AI clouds and AI-optimized colocation in its own Tier III data centers.

- Page: https://gpuclouds.fyi/us/global/online/whitefiber
- Categories: GPU neoclouds, Bare-metal GPU clusters
- Address: Online, Global, United States
- Website: <https://www.whitefiber.com/>
- Updated: 2026-11-07

### [Zenlayer (Distributed Inference / GPU compute)](https://gpuclouds.fyi/us/global/online/zenlayer-distributed-inference-gpu-compute.md)

Zenlayer offers GPU compute for distributed AI inference with NVIDIA RTX 4090, H100 and H200, connected over its global private network spanning Asia, the Middle East, Africa, Europe and the Americas.

- Page: https://gpuclouds.fyi/us/global/online/zenlayer-distributed-inference-gpu-compute
- Categories: Inference platforms, GPU neoclouds
- Address: Online, Global, United States
- Website: <https://www.zenlayer.com/>
- Updated: 2026-11-07

## Online, Global, British Virgin Islands

### [OpenGPU Network](https://gpuclouds.fyi/vg/global/online/opengpu-network.md)

Decentralized blockchain-based GPU network connecting providers and clients for AI inference, training and rendering; offers Relay OpenAI-compatible model API, hourly Relay Instances, client app and reserved capacity.

- Page: https://gpuclouds.fyi/vg/global/online/opengpu-network
- Categories: Decentralized GPU, Inference platforms
- Address: Online, Global, British Virgin Islands
- Website: <https://opengpu.network/>
- Updated: 2026-11-07

## Online, Global, Vietnam

### [FPT AI Factory (FPT Cloud)](https://gpuclouds.fyi/vn/global/online/fpt-ai-factory-fpt-cloud.md)

FPT AI Factory, an NVIDIA Preferred Partner, offers GPU containers, GPU VMs, bare-metal and Kubernetes GPU clusters with B300, H200, H100 and RTX PRO 6000 in Japan and Vietnam, plus serverless and dedicated inference.

- Page: https://gpuclouds.fyi/vn/global/online/fpt-ai-factory-fpt-cloud
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Vietnam
- Website: <https://factory.fpt.ai/>
- Updated: 2026-10-29

### [Viettel IDC (Viettel Cloud GPU)](https://gpuclouds.fyi/vn/global/online/viettel-idc-viettel-cloud-gpu.md)

Viettel IDC's Viettel Cloud GPU provides virtual servers with NVIDIA Tesla T4 GPUs in 16GB and 32GB configurations for rendering, analytics, AI and machine learning, priced monthly in VND.

- Page: https://gpuclouds.fyi/vn/global/online/viettel-idc-viettel-cloud-gpu
- Categories: Sovereign and regional GPU
- Address: Online, Global, Vietnam
- Website: <https://viettel-idc.com.vn/>
- Updated: 2026-11-07

### [VNG Cloud (GPU Cloud)](https://gpuclouds.fyi/vn/global/online/vng-cloud-gpu-cloud.md)

VNG Cloud's GPU Cloud provides on-demand, pay-as-you-go GPU instances with NVIDIA H100, GH200, L40S, A40, RTX 5090 and RTX 4090 for AI/ML, deep learning and rendering, with Kubernetes container management.

- Page: https://gpuclouds.fyi/vn/global/online/vng-cloud-gpu-cloud
- Categories: Sovereign and regional GPU, GPU neoclouds
- Address: Online, Global, Vietnam
- Website: <https://vngcloud.vn/>
- Updated: 2026-11-07

## Academy

- [GPU cloud foundations for buyers](https://gpuclouds.fyi/academy/gpu-cloud-foundations.md): Learn the words on a GPU cloud quote, the main types of provider and how to read a price, so you can read a listing and ask the right questions.
- [Run a GPU cloud trial before you commit](https://gpuclouds.fyi/academy/run-a-gpu-cloud-trial.md): A practical method for choosing a GPU cloud: write a brief and an acceptance rule, run a reproducible trial, compare total cost and put vendor answers in writing.
- [Review a GPU cloud for sensitive workloads](https://gpuclouds.fyi/academy/review-a-gpu-cloud-for-sensitive-workloads.md): A structured review for teams that plan to send prompts, training data or model files to a GPU cloud: data classes, access, tenant boundary, data lifecycle and evidence.

Each course ends with an exam and a certificate of completion. Check a certificate's verification code at `https://gpuclouds.fyi/api/v1/certificates/{code}`.

## GPU buying guides

- [Run inference yourself or buy a managed service](https://gpuclouds.fyi/guides/build-or-buy-inference.md): Compare control, operations work and cost per accepted response.
- [Choose a GPU cloud for your workload](https://gpuclouds.fyi/guides/choose-gpu-cloud.md): Use a repeatable benchmark and a clear acceptance rule.
- [Prepare your first GPU cloud run](https://gpuclouds.fyi/guides/first-gpu-cloud-run.md): Make a small run reproducible, bounded and easy to remove.
- [Compare GPU cloud approaches](https://gpuclouds.fyi/guides/gpu-cloud-approaches.md): Match the service model to control, capacity and operations needs.
- [GPU cloud terms for buyers](https://gpuclouds.fyi/guides/gpu-cloud-glossary.md): Read quotes with clear definitions of capacity and billing.
- [Read a GPU cloud price before you compare it](https://gpuclouds.fyi/guides/gpu-cloud-pricing.md): Compare billing units, GPU variants and extra charges.
- [A GPU cloud security review](https://gpuclouds.fyi/guides/gpu-security-checklist.md): Check data control, access and evidence before upload.
- [Questions to ask a GPU vendor](https://gpuclouds.fyi/guides/vendor-questions.md): Get capacity, billing and exit terms in writing.

## Data sources

- [Listing enrichment](https://gpuclouds.fyi): license source, refreshed manual
- [Vendor websites and documentation](https://gpuclouds.fyi/about): license Facts compiled from public vendor pages. Source text and trademarks remain with their owners., refreshed monthly

## Optional

- [Sitemap](https://gpuclouds.fyi/sitemap.xml): every indexable page