Vast.ai GPU Cloud rents isolated GPU instances across a marketplace of independent hosts, spanning tens of thousands of GPUs in more than 40 data centers. Instances launch through a web console, CLI, Python SDK, or REST API with per-second billing.
Buyers choose on-demand, interruptible, or reserved pricing across NVIDIA architectures from RTX 3090 through B200, with one-click framework templates and dedicated Secure Cloud environments for regulated workloads.

Vast.ai GPU Cloud rents on-demand GPU instances from a distributed host marketplace.


Verified-counterparty venue trading GPU hardware, reserved rental, and forward contracts.


CoreWeave Cloud is an AI-native cloud platform delivering GPU compute for AI workloads.


GPUonCLOUD provides AI and computing teams on-demand GPU instances with instant deployment and hourly pricing.


Lambda Cloud is Lambda's public GPU cloud offering on-demand instances, reserved clusters, and single-tenant supercomputing capacity.


Modal is a high-performance AI infrastructure cloud for online, batch, and multi-modal inference with instant autoscaling.



On-demand NVIDIA GPU rental across certified data centers with per-minute billing.


Pareto Inference spot-rents whole idle 8×NVIDIA B300 nodes to customers until reclaiming them.
Verified-counterparty venue trading GPU hardware, reserved rental, and forward contracts.

CoreWeave Cloud is an AI-native cloud platform delivering GPU compute for AI workloads.

GPUonCLOUD provides AI and computing teams on-demand GPU instances with instant deployment and hourly pricing.

Lambda Cloud is Lambda's public GPU cloud offering on-demand instances, reserved clusters, and single-tenant supercomputing capacity.

Modal is a high-performance AI infrastructure cloud for online, batch, and multi-modal inference with instant autoscaling.

RunPod Cloud is a GPU cloud computing platform for AI and ML workloads.

On-demand NVIDIA GPU rental across certified data centers with per-minute billing.

Pareto Inference spot-rents whole idle 8×NVIDIA B300 nodes to customers until reclaiming them.



Vultr Cloud GPU supplies on-demand and dedicated GPU infrastructure to businesses for accelerated computing.


Scaleway GPU Instances are GPU-based cloud instances operated by Scaleway with data residency in Europe.


Vultr Serverless Inference deploys and serves generative AI models for businesses through managed GPU infrastructure.


Vultr Bare Metal provides dedicated single-tenant servers for high-demand workloads such as AI, analytics and rendering.


GMI Cloud Studio is an AI workflow orchestration and deployment environment for production pipelines on dedicated GPU infrastructure.


Vultr Cloud Compute provides on-demand virtual machines with shared or dedicated vCPUs launched from a control panel or API.



Fastest cloud for training/inference of 1B-24T models with no sharding.




Persistent hosted Linux sandboxes for long-horizon agents, billed per actively-used compute, with pause, resume, and fork.


GMI Cloud Inference is a Model-as-a-Service platform serving hosted LLM, image, video, and audio models through one unified API.


Cloud servers and infrastructure for projects needing flexible, API-manageable computing resources.


DigitalOcean Droplets are on-demand virtual machines with shared or dedicated vCPUs, managed from a control panel or API.


Vultr Cloud GPU supplies on-demand and dedicated GPU infrastructure to businesses for accelerated computing.

Scaleway GPU Instances are GPU-based cloud instances operated by Scaleway with data residency in Europe.

Vultr Serverless Inference deploys and serves generative AI models for businesses through managed GPU infrastructure.

Vultr Bare Metal provides dedicated single-tenant servers for high-demand workloads such as AI, analytics and rendering.

GMI Cloud Studio is an AI workflow orchestration and deployment environment for production pipelines on dedicated GPU infrastructure.

Vultr Cloud Compute provides on-demand virtual machines with shared or dedicated vCPUs launched from a control panel or API.

Managed AI training and development service delivering DGX infrastructure through hyperscale cloud partners.

Fastest cloud for training/inference of 1B-24T models with no sharding.

Replicate is a cloud API to run, fine-tune, and deploy open-source machine learning models with auto-scaling.

GroqCloud is Groq's AI inference platform, structured as GroqMetal, GroqCore, and GroqAssured tiers for infrastructure, inference, and enterprise control.

Persistent hosted Linux sandboxes for long-horizon agents, billed per actively-used compute, with pause, resume, and fork.

GMI Cloud Inference is a Model-as-a-Service platform serving hosted LLM, image, video, and audio models through one unified API.

Cloud servers and infrastructure for projects needing flexible, API-manageable computing resources.

DigitalOcean Droplets are on-demand virtual machines with shared or dedicated vCPUs, managed from a control panel or API.

Containerized inference microservices providing prebuilt APIs for popular AI models on NVIDIA accelerated infrastructure.

Boat provides AI agents persistent Ubuntu sandbox virtual machines with SSH, Docker, and a desktop.
