Page 3 | Best Cloud GPU Providers for Cloud of 2025

Voltage Park

Voltage Park is a next-generation GPU cloud infrastructure provider, offering on-demand and reserved access to NVIDIA HGX H100 GPUs housed in Dell PowerEdge XE9680 servers, each equipped with 1TB of RAM and v52 CPUs. Their six Tier 3+ data centers across the U.S. ensure high availability and reliability, featuring redundant power, cooling, network, fire suppression, and security systems. A state-of-the-art 3200 Gbps InfiniBand network facilitates high-speed communication and low latency between GPUs and workloads. Voltage Park emphasizes uncompromising security and compliance, utilizing Palo Alto firewalls and rigorous protocols, including encryption, access controls, monitoring, disaster recovery planning, penetration testing, and regular audits. With a massive inventory of 24,000 NVIDIA H100 Tensor Core GPUs, Voltage Park enables scalable compute access ranging from 64 to 8,176 GPUs.

Starting Price: $1.99 per hour

View Provider

NVIDIA DGX Cloud Lepton

NVIDIA

NVIDIA DGX Cloud Lepton is an AI platform that connects developers to a global network of GPU compute across multiple cloud providers through a single platform. It offers a unified experience to discover and utilize GPU resources, along with integrated AI services to streamline the deployment lifecycle across multiple clouds. Developers can start building with instant access to NVIDIA’s accelerated APIs, including serverless endpoints, prebuilt NVIDIA Blueprints, and GPU-backed compute. When it’s time to scale, DGX Cloud Lepton powers seamless customization and deployment across a global network of GPU cloud providers. It enables frictionless deployment across any GPU cloud, allowing AI applications to be deployed across multi-cloud and hybrid environments with minimal operational burden, leveraging integrated services for inference, testing, and training workloads.

View Provider

CUDO Compute

CUDO Compute is a high-performance GPU cloud platform built for AI workloads, offering on-demand and reserved clusters designed to scale. Users can deploy powerful GPUs for demanding AI tasks, choosing from a global pool of high-performance GPUs such as NVIDIA H100 SXM, H100 PCIe, HGX B200, GB200 NVL72, A800 PCIe, H200 SXM, B100, A40, L40S, A100 PCIe, V100, RTX 4000 SFF Ada, RTX A4000, RTX A5000, RTX A6000, and AMD MI250/300. It allows spinning up instances in seconds, providing full control to run AI workloads with speed and flexibility to scale globally while meeting compliance requirements. CUDO Compute offers flexible virtual machines for agile workloads, ideal for development, testing, and lightweight production, featuring minute-based billing, high-speed NVMe storage, and full configurability. For teams requiring direct hardware access, dedicated bare metal servers deliver maximum performance without virtualization.

Starting Price: $1.73 per hour

View Provider

AceCloud

AceCloud is a comprehensive public cloud and cybersecurity platform designed to support businesses with scalable, secure, and high-performance infrastructure. Its public cloud services include compute options tailored for RAM-intensive, CPU-intensive, and spot instances, as well as cloud GPU offerings featuring NVIDIA A2, A30, A100, L4, L40S, RTX A6000, RTX 8000, and H100 GPUs. It provides Infrastructure as a Service (IaaS), enabling users to deploy virtual machines, storage, and networking resources on demand. Storage solutions encompass object storage, block storage, volume snapshots, and instance backups, ensuring data integrity and accessibility. AceCloud also offers managed Kubernetes services for container orchestration and supports private cloud deployments, including fully managed cloud, one-time deployment, hosted private cloud, and virtual private servers.

Starting Price: $0.0073 per hour

View Provider

Skyportal

Skyportal is a GPU cloud platform built for AI engineers, offering 50% less cloud costs and 100% GPU performance. It provides a cost-effective GPU infrastructure for machine learning workloads, eliminating unpredictable cloud bills and hidden fees. Skyportal has seamlessly integrated Kubernetes, Slurm, PyTorch, TensorFlow, CUDA, cuDNN, and NVIDIA Drivers, fully optimized for Ubuntu 22.04 LTS and 24.04 LTS, allowing users to focus on innovating and scaling with ease. It offers high-performance NVIDIA H100 and H200 GPUs optimized specifically for ML/AI workloads, with instant scalability and 24/7 expert support from a team that understands ML workflows and optimization. Skyportal's transparent pricing and zero egress fees provide predictable costs for AI infrastructure. Users can share their AI/ML project requirements and goals, deploy models within the infrastructure using familiar tools and frameworks, and scale their infrastructure as needed.

Starting Price: $2.40 per hour

View Provider

Thunder Compute

Thunder Compute is a cloud platform that virtualizes GPUs over TCP, allowing developers to scale from CPU-only machines to GPU clusters with a single command. By tricking computers into thinking they're directly attached to GPUs located elsewhere, Thunder Compute enables CPU-only machines to behave as if they have dedicated GPUs, while the physical GPUs are actually shared among several machines. This approach improves GPU utilization and reduces costs by allowing multiple workloads to run on a single GPU with dynamic memory sharing. Developers can start by building and debugging on a CPU-only machine and then scale to a massive GPU cluster with just one command, eliminating the need for extensive configuration and reducing the costs associated with paying for idle compute resources during development. Thunder Compute offers on-demand access to GPUs like NVIDIA T4, A100 40GB, and A100 80GB, with competitive rates and high-speed networking.

Starting Price: $0.27 per hour

View Provider

Massed Compute

Massed Compute offers high-performance GPU computing solutions tailored for AI, machine learning, scientific simulations, and data analytics. As an NVIDIA Preferred Partner, it provides access to a comprehensive catalog of enterprise-grade NVIDIA GPUs, including A100, H100, L40, and A6000, ensuring optimal performance for various workloads. Users can choose between bare metal servers for maximum control and performance or on-demand compute instances for flexibility and scalability. Massed Compute's Inventory API allows seamless integration of GPU resources into existing business platforms, enabling provisioning, rebooting, and management of instances with ease. Massed Compute's infrastructure is housed in Tier III data centers, offering consistent uptime, advanced redundancy, and efficient cooling systems. With SOC 2 Type II compliance, the platform ensures high standards of security and data protection.

Starting Price: $21.60 per hour

View Provider

NVIDIA Run:ai

NVIDIA

NVIDIA Run:ai is an enterprise platform designed to optimize AI workloads and orchestrate GPU resources efficiently. It dynamically allocates and manages GPU compute across hybrid, multi-cloud, and on-premises environments, maximizing utilization and scaling AI training and inference. The platform offers centralized AI infrastructure management, enabling seamless resource pooling and workload distribution. Built with an API-first approach, Run:ai integrates with major AI frameworks and machine learning tools to support flexible deployment anywhere. It also features a powerful policy engine for strategic resource governance, reducing manual intervention. With proven results like 10x GPU availability and 5x utilization, NVIDIA Run:ai accelerates AI development cycles and boosts ROI.

View Provider

Renderro

With a click of the button open your own high performance PC, on any device, anywhere and anytime. Perform smoothly with up to 96 x 2.8 Ghz, 1360 GB of RAM and 16 x NVIDIA A100 80 GB. Enlarge storage space and computer specs as you need. We keep it simple, so you can focus on what’s really important - your projects. Pick one of our plans, depending  if you want to use the Cloud PC individually or in a team. Decide what hardware setup you want to work with. Work on your Cloud Desktop within your browser or in the desktop app, regardless where you are. Renderro Cloud Storage lets you store all your top-notch designs and resources in a single, easily accessible place. The Cloud Storage is scalable, which means you are not limited by the file size of your projects, and can always manage the storage size at any time. Cloud Drives can be shared between multiple Cloud Desktops, giving you a way to quickly switch between machines, without the need to transfer your media back and forth.

View Provider

Infomaniak

Infomaniak Network

Infomaniak is a major cloud player in Europe and the leading developer of web technologies in Switzerland. From the design of data centers and products to the orchestration of cloud infrastructures, Infomaniak is a Swiss cloud player that controls its value chain from end to end and is exclusively owned by its employees. This independence enables it to guarantee the security, confidentiality and sovereignty of the data of more than one million users in more than 208 countries. At the heart of Europe in Geneva and Winterthur, Infomaniak develops all the solutions that companies need to ensure their online visibility and sustainable development.

View Provider

Rafay

Delight developers and operations teams with the self-service and automation they need, with the right mix of standardization and control that the business requires. Centrally specify and manage configurations (in Git) for clusters encompassing security policy and software add-ons such as service mesh, ingress controllers, monitoring, logging, and backup and restore solutions. Blueprints and add-on lifecycle management can easily be applied to greenfield and brownfield clusters centrally. Blueprints can also be shared across multiple teams for centralized governance of add-ons deployed across the fleet. For environments requiring agile development cycles, users can go from a Git push to an updated application on managed clusters in seconds — 100+ times a day. This is particularly suited for developer environments where updates are very frequent.

View Provider

CoreWeave

CoreWeave is a cloud infrastructure provider specializing in GPU-based compute solutions tailored for AI workloads. The platform offers scalable, high-performance GPU clusters that optimize the training and inference of AI models, making it ideal for industries like machine learning, visual effects (VFX), and high-performance computing (HPC). CoreWeave provides flexible storage, networking, and managed services to support AI-driven businesses, with a focus on reliability, cost efficiency, and enterprise-grade security. The platform is used by AI labs, research organizations, and businesses to accelerate their AI innovations.

View Provider

NVIDIA DGX Cloud

NVIDIA

NVIDIA DGX Cloud offers a fully managed, end-to-end AI platform that leverages the power of NVIDIA’s advanced hardware and cloud computing services. This platform allows businesses and organizations to scale AI workloads seamlessly, providing tools for machine learning, deep learning, and high-performance computing (HPC). DGX Cloud integrates seamlessly with leading cloud providers, delivering the performance and flexibility required to handle the most demanding AI applications. This service is ideal for businesses looking to enhance their AI capabilities without the need to manage physical infrastructure.

View Provider

IBM GPU Cloud Server

IBM

We listened and lowered our bare metal and virtual server prices. Same power and flexibility. A graphics processing unit (GPU) is “extra brain power” the CPU lacks. Choosing IBM Cloud® for your GPU requirements gives you direct access to one of the most flexible server-selection processes in the industry, seamless integration with your IBM Cloud architecture, APIs and applications, and a globally distributed network of data centers. IBM Cloud Bare Metal Servers with GPUs perform better on 5 TensorFlow ML models than AWS servers. We offer bare metal GPUs and virtual server GPUs. Google Cloud only offers virtual server instances. Like Google Cloud, Alibaba Cloud only offers GPU options on virtual machines.

View Provider

Genesis Cloud

Whether you're creating machine learning models or conducting complex data analytics, Genesis Cloud provides the accelerators for any size application. Create a GPU or CPU virtual machine in minutes. With multiple configurations, you will find an option that works for your project's size, from bootstrap to scaleout. Create storage volumes that can dynamically expand as your data grows. Backed by a highly available storage cluster and encrypted at rest, your data is secure from unexpected loss or access. Our data centers are built using a non-blocking leaf-spine architecture based on 100G switches. Each server is connected with multiple 25G uplinks and each account has its own isolated virtual network for added privacy and security. Our cloud offers you infrastructure powered by renewable energy at a price that is the most affordable in the market.

View Provider

Vast.ai

Vast.ai is the market leader in low-cost cloud GPU rental. Use one simple interface to save 5-6X on GPU compute. Use on-demand rentals for convenience and consistent pricing. Or save a further 50% or more with interruptible instances using spot auction based pricing. Vast has an array of providers that offer different levels of security: from hobbyists up to Tier-4 data centers. Vast.ai helps you find the best pricing for the level of security and reliability you need. Use our command line interface to search the entire marketplace for offers while utilizing scriptable filters and sort options. Launch instances quickly right from the CLI and easily automate your deployment. Save an additional 50% or more by using interruptible instances and auction pricing. The highest bidding instances run; other conflicting instances are stopped.

Starting Price: $0.20 per hour

View Provider

HOSTKEY

We prioritize your budget, therefore by choosing our service, you’re guaranteed to get the assistance that meets your expectations but doesn’t exceed resources. We offer an agile and flexible product customized to your requests. Every client gets a maximally customized treatment. We are ready to meet your most advanced needs in server configuration. Every server we sell has been assembled and checked personally. Qualified personnel, and professional services for both the experienced and the newbies. However complicated, a project never scares us. We have earned our clients’ respect and built a reputation to be proud of. We speak the language of IT specialists from sales to day-to-day support, resellers and affiliates enjoy superior conditions, we call back with regular sales and special offers.

Starting Price: €60 per month

View Provider

DataCrunch

Up to 8 NVidia® H100 80GB GPUs, each containing 16896 CUDA cores and 528 Tensor Cores. This is the current flagship silicon from NVidia®, unbeaten in raw performance for AI operations. We deploy the SXM5 NVLINK module, which offers a memory bandwidth of 2.6 Gbps and up to 900GB/s P2P bandwidth. Fourth generation AMD Genoa, up to 384 threads with a boost clock of 3.7GHz. We only use the SXM4 'for NVLINK' module, which offers a memory bandwidth of over 2TB/s and Up to 600GB/s P2P bandwidth. Second generation AMD EPYC Rome, up to 192 threads with a boost clock of 3.3GHz. The name 8A100.176V is composed as follows: 8x RTX A100, 176 CPU core threads & virtualized. Despite having less tensor cores than the V100, it is able to process tensor operations faster due to a different architecture. Second generation AMD EPYC Rome, up to 96 threads with a boost clock of 3.35GHz.

Starting Price: $3.01 per hour

View Provider

Cirrascale

Our high-throughput storage systems can serve millions of small, random files to GPU-based training servers accelerating overall training times. We offer high-bandwidth, low-latency networks for connecting distributed training servers as well as transporting data between storage and servers. Other cloud providers squeeze you with extra fees and charges to get your data out of their storage clouds, and those can add up fast. We consider ourselves an extension of your team. We work with you to set up scheduling services, help with best practices, and provide superior support. Workflows can vary from company to company. Cirrascale works to ensure you get the right solution for your needs to get you the best results. Cirrascale is the only provider that works with you to tailor your cloud instances to increase performance, remove bottlenecks, and optimize your workflow. Cloud-based solutions to accelerate your training, simulation, and re-simulation time.

Starting Price: $2.49 per hour

View Provider

TensorDock

All products come with bandwidth included and are usually between 70 to 90% cheaper than competing products on the market. They're developed in-house by our 100% US-based team. Servers are operated by independent hosts that run our hypervisor software. Flexible, resilient, scalable, and secure cloud for burstable workloads. Up to 70% cheaper than incumbent clouds. Low-cost secure servers on monthly or longer terms for continuous workloads (e.g. ML inference). Being integrated with our customers' tech stacks is a focus of our business. Well-documented, well-maintained, well-everything.

Starting Price: $0.05 per hour

View Provider

Together AI

Whether prompt engineering, fine-tuning, or training, we are ready to meet your business demands. Easily integrate your new model into your production application using the Together Inference API. With the fastest performance available and elastic scaling, Together AI is built to scale with your needs as you grow. Inspect how models are trained and what data is used to increase accuracy and minimize risks. You own the model you fine-tune, not your cloud provider. Change providers for whatever reason, including price changes. Maintain complete data privacy by storing data locally or in our secure cloud.

Starting Price: $0.0001 per 1k tokens

View Provider

Lease Packet

Lease Packet is managed server provider. We have all types of servers that can be further customized as per your requirements. Find the best dedicated servers, VPS servers, Cloud servers, GPU servers, Colocation servers, Streaming servers, 10 GBPS servers, Mass mailing servers, storage servers etc. all in one place. Our startup servers, enterprises servers and sharks servers make sure businesses of all size can benefits from our services. Additionally, we can help you with your AWS billing optimization by becoming your AWS billing partner. We make sure all your AWS resources are utilized in the right place to offer your maximum efficiency. All our managed servers come with a 99% uptime guarantee and 24x7 server support for instant resolution. Whether you're a startup, an established enterprise, or an individual with a passion project, we have the expertise and resources to support your goals. Visit our website and learn more about our server solutions.

Starting Price: $10

View Provider

Node AI

Spend less time and money on infrastructure and more time on your business. Get more value from your GPU investment. Our platform is where complexity meets simplicity, providing a seamless interface for clients to tap into a global network of AI nodes. Clients submit their computational tasks to Node AI, where they are instantly distributed across our secure network of high-performance AI nodes. The tasks are processed in parallel, harnessing the power of the L1 Blockchain for secure, efficient, and verifiable computation. Verified results are encrypted and returned to the clients promptly, ensuring confidentiality and integrity.

View Provider

Runyour AI

From renting machines for AI research to specialized templates and servers, Runyour AI provides the optimal environment for artificial intelligence research. Runyour AI is an AI cloud service that provides easy access to GPU resources and research environments for artificial intelligence research. You can rent various high-performance GPU machines and environments at a reasonable price. Additionally, you can register your own GPUs to generate revenue. Transparent billing policy where you pay for charging points used through minute-by-minute real-time monitoring. From casual hobbyists to seasoned researchers, we provide specialized GPUs for AI projects, catering to a range of needs. An AI project environment that is easy and convenient for even first-time users. By utilizing Runyour AI's GPU machines, you can kickstart your AI research with minimal setup. Designed for quick access to GPUs, it provides a seamless research environment for machine learning and AI development.

View Provider

Burncloud

Burncloud is a leading cloud computing service provider focused on delivering efficient, reliable, and secure GPU rental solutions for businesses. Our platform operates on a systemized model designed to meet the high-performance computing needs of various enterprises. Core Services Online GPU Rental Services: We offer a variety of GPU models for rent, including data center-grade devices and edge consumer-level computing equipment, to meet the diverse computational needs of businesses. Our best-selling products currently include: RTX 4070, RTX 3070 Ti, H100 PCIe, RTX 3090 Ti, RTX 3060, NVIDIA 4090, L40, RTX 3080 Ti, L40S, RTX 4090, RTX 3090, A10, H100 SXM, H100 NVL, A100 PCIe 80GB, and more. Compute Cluster Setup Services: Our technical team has extensive experience in IB networking technology and has successfully completed the setup of five 256-node clusters. For cluster setup services, please contact the customer service team on the Burncloud official website.

Starting Price: $0.03/hour

View Provider

Amazon EC2 P5 Instances

Amazon

Amazon Elastic Compute Cloud (Amazon EC2) P5 instances, powered by NVIDIA H100 Tensor Core GPUs, and P5e and P5en instances powered by NVIDIA H200 Tensor Core GPUs deliver the highest performance in Amazon EC2 for deep learning and high-performance computing applications. They help you accelerate your time to solution by up to 4x compared to previous-generation GPU-based EC2 instances, and reduce the cost to train ML models by up to 40%. These instances help you iterate on your solutions at a faster pace and get to market more quickly. You can use P5, P5e, and P5en instances for training and deploying increasingly complex large language models and diffusion models powering the most demanding generative artificial intelligence applications. These applications include question-answering, code generation, video and image generation, and speech recognition. You can also use these instances to deploy demanding HPC applications at scale for pharmaceutical discovery.

View Provider

Amazon EC2 Capacity Blocks for ML

Amazon

Amazon EC2 Capacity Blocks for ML enable you to reserve accelerated compute instances in Amazon EC2 UltraClusters for your machine learning workloads. This service supports Amazon EC2 P5en, P5e, P5, and P4d instances, powered by NVIDIA H200, H100, and A100 Tensor Core GPUs, respectively, as well as Trn2 and Trn1 instances powered by AWS Trainium. You can reserve these instances for up to six months in cluster sizes ranging from one to 64 instances (512 GPUs or 1,024 Trainium chips), providing flexibility for various ML workloads. Reservations can be made up to eight weeks in advance. By colocating in Amazon EC2 UltraClusters, Capacity Blocks offer low-latency, high-throughput network connectivity, facilitating efficient distributed training. This setup ensures predictable access to high-performance computing resources, allowing you to plan ML development confidently, run experiments, build prototypes, and accommodate future surges in demand for ML applications.

View Provider

Amazon EC2 UltraClusters

Amazon

Amazon EC2 UltraClusters enable you to scale to thousands of GPUs or purpose-built machine learning accelerators, such as AWS Trainium, providing on-demand access to supercomputing-class performance. They democratize supercomputing for ML, generative AI, and high-performance computing developers through a simple pay-as-you-go model without setup or maintenance costs. UltraClusters consist of thousands of accelerated EC2 instances co-located in a given AWS Availability Zone, interconnected using Elastic Fabric Adapter (EFA) networking in a petabit-scale nonblocking network. This architecture offers high-performance networking and access to Amazon FSx for Lustre, a fully managed shared storage built on a high-performance parallel file system, enabling rapid processing of massive datasets with sub-millisecond latencies. EC2 UltraClusters provide scale-out capabilities for distributed ML training and tightly coupled HPC workloads, reducing training times.

View Provider

AWS Elastic Fabric Adapter (EFA)

United States

Elastic Fabric Adapter (EFA) is a network interface for Amazon EC2 instances that enables customers to run applications requiring high levels of inter-node communications at scale on AWS. Its custom-built operating system (OS) bypass hardware interface enhances the performance of inter-instance communications, which is critical to scaling these applications. With EFA, High-Performance Computing (HPC) applications using the Message Passing Interface (MPI) and Machine Learning (ML) applications using NVIDIA Collective Communications Library (NCCL) can scale to thousands of CPUs or GPUs. As a result, you get the application performance of on-premises HPC clusters with the on-demand elasticity and flexibility of the AWS cloud. EFA is available as an optional EC2 networking feature that you can enable on any supported EC2 instance at no additional cost. Plus, it works with the most commonly used interfaces, APIs, and libraries for inter-node communications.

View Provider

CoresHub

Coreshub provides GPU cloud services, AI training clusters, parallel file storage, and image repositories, delivering secure, reliable, and high-performance cloud-based AI training and inference environments. The platform offers a range of solutions, including computing power market, model inference, and various industry-specific applications. Coreshub's core team comprises experts from Tsinghua University, leading AI companies, IBM, renowned venture capital firms, and major internet corporations, bringing extensive AI technical expertise and ecosystem resources. The platform emphasizes an independent and open cooperative ecosystem, actively collaborating with AI model suppliers and hardware manufacturers. Coreshub's AI computing platform enables unified scheduling and intelligent management of diverse heterogeneous computing power, meeting AI computing operation, maintenance, and management needs in a one-stop manner.

Starting Price: $0.24 per hour

View Provider

Best Cloud GPU Providers for Cloud - Page 3

Compare the Top Cloud GPU Providers for Cloud as of August 2025 - Page 3

Voltage Park

NVIDIA DGX Cloud Lepton

CUDO Compute

AceCloud

Skyportal

Thunder Compute

Massed Compute

NVIDIA Run:ai

Renderro

Infomaniak

Rafay

CoreWeave

NVIDIA DGX Cloud

IBM GPU Cloud Server

Genesis Cloud

Vast.ai

HOSTKEY

DataCrunch

Cirrascale

TensorDock

Together AI

Lease Packet

Node AI

Runyour AI

Burncloud

Amazon EC2 P5 Instances

Amazon EC2 Capacity Blocks for ML

Amazon EC2 UltraClusters

AWS Elastic Fabric Adapter (EFA)

CoresHub

Best Cloud GPU Providers for Cloud - Page 3

Compare the Top Cloud GPU Providers for Cloud as of August 2025 - Page 3

Voltage Park

NVIDIA DGX Cloud Lepton

CUDO Compute

AceCloud

Skyportal

Thunder Compute

Massed Compute

NVIDIA Run:ai

Renderro

Infomaniak

Rafay

CoreWeave

NVIDIA DGX Cloud

IBM GPU Cloud Server

Genesis Cloud

Vast.ai

HOSTKEY

DataCrunch

Cirrascale

TensorDock

Together AI

Lease Packet

Node AI

Runyour AI

Burncloud

Amazon EC2 P5 Instances

Amazon EC2 Capacity Blocks for ML

Amazon EC2 UltraClusters

AWS Elastic Fabric Adapter (EFA)

CoresHub

Related Categories