Spin up. Scale. Snapshot.
Provision compute across AWS, Azure, GCP, Linode, Hetzner, OVHcloud, Alibaba, and Oracle with one CLI, one Terraform provider, and one REST API. Sub-15-second boots, signed images, and snapshots that survive whole-region failures.
One control plane — every cloud you actually use
Instance families
From $2/mo dev boxes to 96-vCPU bare metal
Same API, same SSH, same snapshot semantics across every family. Pick one — see specs, price, and a recommended workload instantly.
General family
g3.medium
Best for
web apps, app servers, microservices
25+ regions · 4 continents
Pin compute to the region that matches your users
Data residency for regulated workloads, lowest-latency for the rest. Snapshots replicate cross-region by policy so you can survive an entire AZ — or an entire provider — going dark.
- GDPR-compliant EU regions with data pinned at rest and in transit
- HIPAA + PCI scope on US-East and US-West with BAA available
- APAC coverage in Tokyo, Singapore, Mumbai, and Sydney
- Cross-region snapshots — automatic, encrypted, lifecycle-managed
CLI · Terraform · REST
Same shape across every cloud — no provider-specific quirks
Choose the surface that fits the moment: a quick CLI to spin up a one-off, Terraform for durable infrastructure, REST for in-app provisioning. Drift detection and signed images are on by default for all three.
- Idempotent operations — re-run the same command safely
- Drift detection across cloud accounts in one dashboard
- Signed cloud-init scripts and image attestations on every boot
- Snapshots, resize, and migration without losing the data volume
1# Provision a VM in any cloud, in any region — same flags everywhere.2vxcli vm create vm-api-01 \3 --cloud aws \4 --region us-east-1 \5 --type g3.medium \6 --image ubuntu-22.04 \7 --disk 160 \8 --ssh-key team-prod \9 --userdata ./cloud-init.yml10 11# → Provisioning… 13.7s12# → ssh [email protected]13# → cost: $0.084/hr · $60.48/mo14 15# Snapshot before risky migration16vxcli vm snapshot vm-api-01 --label "pre-migration-2026-05-10"17 18# Resize without losing data19vxcli vm resize vm-api-01 --type c5.xlarge --downtime-ok20 21# Burst into spot capacity for batch22vxcli vm create-batch ml-train --type gpu.l4 --count 12 --spot --max-price 0.40Pick the right size from a goal
vxcli vm suggest "Postgres 14, 200GB hot data, 1k QPS" — recommends r5.2xlarge with disk + alerts.
Production-grade by default
Everything a production VM should have on day one
Sub-15s provisioning
Pre-warmed image cache and parallel cloud-init steps cut median boot to under fifteen seconds across all clouds.
Snapshots + lifecycle
Daily snapshots, retention windows, cross-region copies, and one-click restore — set once at the policy level.
Signed images
Every base image is signed, scanned, and SBOM-attested. Cloud-init scripts verify signatures before exec.
Online resize
Resize CPU, memory, or disk without losing the data volume. Most resizes complete in under a minute, no rebuild.
Identity-aware SSH
No shared SSH keys. Tie sessions to your IdP, MFA on access, full keystroke audit log streamed to your SIEM.
Built-in metrics
CPU, memory, disk, network, and load shipped automatically — Prometheus-compatible scrape endpoint included.
VPC + private routes
Per-tenant VPC, private subnets, security groups locked-down by default, optional WireGuard mesh between sites.
GPU + bare-metal
L4, A10, and H100 GPUs available on-demand and via spot. EPYC and Xeon bare metal for sovereign workloads.
One operator surface
Console, metrics, snapshots, billing, and IAM in one dashboard — no cloud-specific tab-switching.
Real workloads, measured
Production traffic — not synthetic benchmarks
Aggregated from 30 days of customer provisioning. Median across all 8 cloud providers, all 25+ regions, all 6 instance families.
What teams run on these VMs
From one-off dev boxes to bare-metal database fleets
Application & API hosting
Run Node, FastAPI, Rails, or Go services with SSH, autoscaling, and a managed load balancer in front — same shape on any cloud.
Self-managed databases
Postgres, MySQL, and Mongo on memory-optimized instances with daily snapshots, cross-region copies, and instant clone-for-staging.
GPU inference fleet
L4 and A10 GPUs on-demand or spot, autoscaled by queue depth. Inference at $0.96/hr, no separate GPU cluster to operate.
CI runners & batch jobs
Compute-optimized fleets that scale to hundreds of runners during deploys and burn down to zero on quiet weekends.
Sovereign / regulated workloads
Bare-metal in EU regions with FIPS-eligible kernels and air-gapped controls — meet GDPR, PCI, and HIPAA without changing tooling.
Edge & branch deployments
Lightweight VMs in Hetzner, OVH, Linode, and Alibaba edge regions, with WireGuard mesh tunnels back to the core platform.
One CLI. Eight clouds. Zero lock-in.
Provision your first VM in under fifteen seconds — across the cloud and region of your choice, with snapshots and identity-aware SSH on by default.