Vultr is unclaimed —

Vultr

The AI-first Global Cloud Platform·vultr.com

Vultr is a global cloud hosting platform offering SSD VPS cloud servers, bare metal servers, GPU cloud instances, managed databases, storage, and Kubernetes across 33 data center regions. It provides on-demand, hourly-billed infrastructure with 100% KVM virtualization for developers and businesses. The platform includes a Marketplace of one-click applications, networking features like VPC and load balancers, and AI/HPC-focused GPU compute powered by NVIDIA and AMD hardware.

What it's for

low traffic websites, blogs, CMSdev/test environmentssmall databasescloud-native and enterprise workloadsweb and application serverse-commercegame serversvideo and audio streamingAPI servingrelational databasesAI training, large-scale inference, and agentic AIHPC deployments

Features 21

Automated backups and manual snapshots
Bare Metal single-tenant dedicated servers
Block Storage
CDN with push and pull zones
Cloud Compute virtual machines (Regular Performance, High Performance, High Frequency, VX1)
Cloud GPU instances with NVIDIA and AMD GPUs
Container Registry
Direct Connect private connectivity
File System (NVMe) storage
IAM (Identity and Access Management)
Load Balancers
Managed Databases for MySQL, PostgreSQL, Apache Kafka, and Valkey
Object Storage (S3-compatible)
One-Click Marketplace app deployment
Optimized Cloud Compute with dedicated AMD EPYC vCPUs
Serverless Inference for AI model deployment
Upload ISO / custom operating systems
Virtual Private Cloud (VPC)
Vultr Agent AI-powered assistant for infrastructure setup
Vultr Firewall and native DDoS Protection
Vultr Kubernetes Engine (managed Kubernetes)

At a glance

free trialAPI

DeploymentCloud

PlatformsWeb

Fordevelopers, businesses

CompanyThe Constant Company, LLC · 2014

Integrations

TerraformDocker HubGitHubGoogle

Complianceself-reported

SOC 2+ (HIPAA)PCI DSSCSA Star Level 1ISO/IEC 20000-1:2018ISO/IEC 27001:2022ISO/IEC 27017:2015ISO/IEC 27018:2019GDPR

Resources

Docs

Pricing

Vultr bills primarily by the hour with no long-term contracts ('Pay only for what you use'). Many products (VX1, Cloud GPU) are priced per hour or per GPU-hour rather than flat monthly.

Cloud Compute (Regular Performance)$2.5/mo

Also billed hourly at $0.004/hr; sizes range up to $640.00/mo (24 vCPUs) in the visible pricing table.

Smallest instance: 1 vCPU, 0.5 GB Memory, 10 GB Storage, 0.5 TB Bandwidth (IPv6 only)

VX1 Cloud ComputeQuote

Positioned for cloud-native and enterprise workloads.

Billed on actual hours used each month

Optimized Cloud Compute$28/mo

Homepage states 'Starting at $28.00/month'; detailed pricing table lists General Purpose tier starting at $30.00/mo (1 vCPU, 4 GB) — figures not fully reconciled in source pages.

Runs on fully dedicated, new-generation AMD EPYC vCPUs with built-in NVMe SSD

Bare Metal$120/mo

Fully automated dedicated single-tenant servers, zero virtualization layer

NVIDIA Cloud GPUQuote

Includes GB300 NVL72 and HGX B300, listed as 'Reserve now' with no stated price.

AMD Cloud GPUQuote

AMD Instinct MI455X listed as 'Reserve now' with no stated price.