Marketplace GPU selection
Choose among 45 listed GPU models, including H100 SXM5 80GB, A100 variants, L40, RTX 4090, RTX 3090, RTX 6000 Ada, RTX A6000, and RTX A4000. Prices vary by host and location.
TensorDock
Freemium
TensorDock is a marketplace-based cloud infrastructure platform for on-demand GPU and CPU servers. It supports machine learning, rendering, cloud gaming, scientific computing, and other workloads that need configurable virtual machines.
What is TensorDock?
TensorDock is a marketplace-based cloud infrastructure platform for on-demand GPU and CPU servers. It provides configurable virtual machines with dedicated GPUs, root access, selectable CPU, RAM, and NVMe storage, and support for customer-managed operating systems and drivers.
The platform is designed for machine learning, AI inference, rendering, cloud gaming, scientific computing, transcoding, and batch processing. Its GPU catalog includes 45 models, ranging from consumer cards such as the RTX 4090 and RTX 3090 to enterprise hardware such as the H100 SXM5 and A100 SXM4. CPU instances use Intel Xeon and AMD EPYC processors.
Customers sign up, deposit funds, and deploy a server on a pay-as-you-go basis. GPU pricing is separate from configurable resources, and rates vary by host and location. TensorDock states that Docker is included on all VM templates and that Windows 10 is supported.
What can TensorDock do?
Configurable virtual machines
Deploy virtual machines with dedicated GPUs, root access, customer-managed drivers, and separately selected CPU, RAM, and block NVMe storage.
Pay-as-you-go billing
Deposit funds before deployment and have the balance deducted continuously while servers run. When the balance reaches zero, deployed servers are automatically deleted.
Global capacity marketplace
TensorDock reports hundreds of GPUs available through its dashboard across more than 100 locations in over 20 countries, with a partner network of up to 30,000 GPUs.
API and VM tooling
The platform provides an API with server metadata and availability information. Docker is included on all VM templates, and Windows 10 support is available for compatible workflows.
Host and access controls
TensorDock says hosts are vetted for hardware and technical capability, host SSH access is revoked, access is restricted to authorized personnel, and hostnodes are monitored for suspicious logins.
Use Cases
“Machine-learning training and inference”
Use dedicated GPUs such as the H100, A100, or consumer RTX models to train models, run inference, or research GAN-based systems with selectable infrastructure resources.
“3D rendering and animation”
GPU instances can provide compute for image processing, animation, and rendering workflows, while Windows support can accommodate software environments that require that operating system.
“Cloud gaming and Windows applications”
Deploy Windows virtual machines for cloud gaming or other interactive GPU workloads. The site cites airgpu as an example of using the TensorDock API to provision Windows VMs.
“Scientific and batch computing”
Use CPU instances built on Intel Xeon or AMD EPYC processors for scientific computing, transcoding, and batch-processing tasks that do not require a GPU.
Frequently Asked Questions
How do I start using TensorDock?
Create an account, deposit funds, and deploy a GPU or CPU server from the dashboard. The site says users can start with a $5 deposit.
How does TensorDock billing work?
TensorDock uses pay-as-you-go billing. Funds are deposited in advance, the balance is deducted continuously after a server is deployed, and servers are automatically deleted when the balance reaches zero. Long-term reserved pricing is available by contacting TensorDock.
Why can the same GPU or CPU have different prices?
TensorDock is a marketplace of independent hosts. Hosts set their own rates, and prices can differ because of location, redundancy, and other host-specific factors. The dashboard should be checked for current pricing and availability.
How does TensorDock address host access to customer data?
TensorDock says most platform hardware is hosted in certified data centers, host SSH access is revoked, access is limited to personnel who need it, and an agent monitors hostnodes for logins and suspicious activity. Customers should review the platform's current security documentation for full details.
Quick Facts
- Product type
- Marketplace-based GPU and CPU cloud infrastructure
- Primary workloads
- Machine learning, AI inference, rendering, cloud gaming, scientific computing, transcoding, and batch processing
- GPU catalog
- 45 GPU models, including H100 SXM5, A100, RTX 4090, and RTX 3090
- Deployment model
- On-demand virtual machines with dedicated GPUs and root access
- Billing model
- Pay as you go; GPU and configurable CPU, RAM, and storage charges are listed separately
- Integrations and tooling
- Docker on all VM templates, Windows 10 support, and an API with metadata and availability information
TensorDock Alternatives
Salad
salad.com
Salad is a distributed GPU cloud for AI, offering usage-based pricing with no contracts or pre-payment for inference, transcription, image generation, and batch processing.
Hyperstack
www.hyperstack.cloud
Hyperstack is a cloud GPU platform for running AI and machine learning workloads, including training, inference, data analytics, and model development. It also provides AI Studio, virtual machines, and managed Kubernetes for deploying and operating GPU-backed workloads.
Radiant
radiant.co
Radiant is an integrated AI infrastructure platform that finances, builds, and operates data centers, GPU systems, networking, storage, and managed services. It helps AI teams and infrastructure operators provision and run compute through a unified platform and FlightDeck control plane.
Together AI
together.ai
Together AI is an AI cloud platform for inference, fine-tuning, GPU clusters, sandboxes, and managed storage.
Paperspace
www.paperspace.com
Paperspace is a cloud platform for developing, training, and deploying machine learning applications with managed notebooks, GPU machines, and deployment workflows. It serves ML developers, data scientists, researchers, and teams that need on-demand accelerated computing.
DeepInfra
deepinfra.com
DeepInfra provides hosted machine-learning model inference and on-demand GPU instances for developers and teams. Its catalog covers text, image, audio, video, embedding, reranking, and other model workloads with pay-as-you-go pricing.