Infrastructure that breathes with you.

Elastic vCPU and RAM that scale in seconds, self-healing failover across 9 regions, and a pay-as-you-scale bill that never punishes success.

Architecture

A stack that heals itself.

Every layer is redundant, monitored, and automatically remediated. No pagers, no on-call heroics.

Anycast Edge + CDN

240+ PoPs · HTTP/3 · Brotli

Healthy

Global Load Balancer

Weighted DNS · Health-check based failover · 30s TTL

Healthy

Compute Cluster (Kubernetes)

Auto-scale groups · Live migration · Zero-downtime updates

Healthy

Distributed NVMe Storage

Ceph-based · 3× replication · Snapshot every 6h

Healthy

Zero-Trust Network

Per-workload identities · mTLS · WAF at edge

Healthy

Global regions

Deploy closest to your users.

Nine regions, one control plane. Move workloads between regions with two clicks — no reprovision.

$ deploy --region eu-west

us-east
Ashburn, USA

Ping
8ms

us-west
Oregon, USA

Ping
14ms

eu-west
Amsterdam, NL

Ping
6ms

eu-central
Frankfurt, DE

Ping
11ms

uk-south
London, UK

Ping
9ms

me-west
Dubai, UAE

Ping
22ms

ap-south
Mumbai, IN

Ping
18ms

ap-east
Singapore, SG

Ping
16ms

sa-east
São Paulo, BR

Ping
28ms

Elastic scaling

Scale up in 12 seconds. Down just as fast.

Set a floor and a ceiling. Our orchestrator watches CPU, memory and request queue depth — and adds vCPUs before your users feel a hitch. When traffic settles, resources retract and your bill goes back down. That’s the whole trick.

Scale-up

12s

Billing

per-min

Max nodes

128

Everything included

No hidden add-ons.

1-click scaling

Vertical + horizontal from the console or API.

Snapshots every 6h

Point-in-time recovery, 30-day retention.

Free DDoS shield

L3-7 protection up to 100 Gbps included.

Managed databases

Postgres, MySQL, Redis — one click to deploy.

Real-time metrics

CPU, RAM, I/O, req/s streamed to your dashboard.

Private networking

VPC peering between clusters at no cost.

Calculator

Price your workload in 10 seconds.

Per-minute billing. Only pay for what you actually use.

vCPU cores

4 cores

RAM

8 GB

NVMe Storage

100 GB

ESTIMATED MONTHLY

$ 58.00 /mo

≈ $ 0.081 /hour · per-minute billed

4 vCPU $ 24.00

8 GB RAM $ 24.00

100 GB NVMe $ 10.00

Launch this config →

Pricing

Or pick a preset and go.

All plans include auto-scaling, snapshots, and 100 GB bandwidth.

Side projects

Cloud Start

$ 6

.99 /mo

Popular

Startups

Cloud Grow

$ 19

.99 /mo

Production

Cloud Scale

$ 44

.99 /mo

High-traffic

Cloud Pro

$ 89

.99 /mo

SUPPORT

Frequently Asked Questions

How does auto-scaling actually work?

You set a floor and a ceiling for vCPU and RAM. Our orchestrator watches CPU, memory, and request-queue depth every 5 seconds and adds capacity in ~12s when thresholds trip. When traffic drops for 5 minutes, resources retract. You’re billed per-minute for whatever’s actually running.

Is my data replicated across regions?

Within a region, all data is 3× replicated on Ceph across independent racks. Cross-region replication is opt-in per volume — enable it in one click for disaster-recovery scenarios. Snapshots are always stored in a separate availability zone.

What happens if a node fails?

Container-level failover is automatic and typically completes in under 30 seconds. Your public IP moves to a healthy node; live TCP connections re-establish on retry. Node-level failure does not require any action from you — we page ourselves, not you.

Can I move a workload between regions?

Yes. From the console or API, select a new region and click Move. We snapshot, ship, and re-launch — typical downtime is 30–90 seconds for a 100 GB workload. No reprovisioning, no config changes.

How is pricing calculated?

Per-minute for compute + storage. There are no ingress fees, no free-tier gotchas, and egress within the same region is free. Cross-region and internet egress are $0.01/GB after the first 100 GB per plan.

Do you offer managed databases?

Yes — Postgres, MySQL, and Redis with automated backups, PITR, and zero-downtime patching. Deploy from the console with one click and connect via a private VPC endpoint.

Deploy in one region. Or nine.

$100 in credits when you launch your first cluster. No card required to start a trial.