All systems operational 14 regions 99.998% uptime (90d)
v3.4.2 shipped Apr 24 — read changelog
Independent · European · Engineer-first

Edge compute,
without the
latency tax.

Deploy in 14 European and Asian regions over our own anycast network. Per-second billing. No egress fees. The console takes you four clicks to a running node.

No credit card · €10 in trial credit · GDPR & ISO 27001
arv · zsh
14
Regions across EU & APAC
12s
Median provision time, p50
99.998%
Uptime, trailing 90 days
0.004
Per vCPU / hour, all regions
Trusted by builders at
Lumen Foundry Trailhead.io numerix.dev PEAKSIDE linkup co. Cardinal Data tidewave overstack
Platform

Six primitives.
One opinionated platform.

No surprise managed services bolted on. Each Arveon product is built around the same control plane, billed by the second, and exposed through one consistent API.

From git push to global

Three steps. No ceremony.

01

Bring your code

Connect a Git repo, push a Docker image, or upload a binary. We don't care how it got built — we care that it runs.

02

Pick a region

14 to choose from. Every region is a real metro with real peering — no "edge locations" that are actually 400 km away.

03

Ship

12 seconds, p50. Healthchecks, IPv6, anycast, certificates — wired up by the time your terminal scrolls.

One platform · three syntaxes

Talk to Arveon however you think.

The CLI, our Terraform provider, and the REST API are generated from the same OpenAPI spec. Pick the one that fits your week.

Provision a 4-vCPU node in eu-hel-1
# install once
$ curl -fsSL https://arveon.space/install.sh | sh

# auth
$ arv auth login

# provision
$ arv compute create \
    --image  ubuntu-24.04 \
    --plan   a4 \
    --region eu-hel-1 \
    --ssh    ~/.ssh/id_ed25519.pub \
    --name   prod-api-01

# done in 12 seconds
prod-api-01  185.4.118.92  eu-hel-1a  running
terraform {
  required_providers {
    arveon = {
      source  = "arveon/arveon"
      version = "~> 0.9"
    }
  }
}

resource "arveon_compute" "api" {
  name   = "prod-api-01"
  image  = "ubuntu-24.04"
  plan   = "a4"
  region = "eu-hel-1"

  network {
    ipv6_only = false
    private_id = arveon_network.core.id
  }
}
$ curl https://api.arveon.space/v1/compute \
    -H "Authorization: Bearer $ARVEON_TOKEN" \
    -H "Content-Type: application/json" \
    -d '{
      "name":   "prod-api-01",
      "image":  "ubuntu-24.04",
      "plan":   "a4",
      "region": "eu-hel-1"
    }'

# 201 Created
{
  "id":     "cmp_4f7c2b9a",
  "status": "provisioning",
  "ipv4":   "185.4.118.92",
  "ipv6":   "2a0c:5240::4f12"
}
Where we live

A real network in real places.

Every region is a metro with carrier-neutral peering. We don't count edge POPs as regions. We don't run regions out of someone else's datacenter. We open them, we own the ASN, we run the BGP.

  • eu-hel-1 · Helsinki
    HQ region · all services
    since 2022
  • eu-fra-1 · Frankfurt
    DE-CIX peering · all services
    since 2022
  • eu-ams-1 · Amsterdam
    AMS-IX · all services
    since 2023
  • eu-par-1 · Paris
    degraded · functions only
    since 2024
  • ap-tyo-1 · Tokyo
    JPNAP · all services
    since 2024
  • All 14 regions
Operators talk

What teams say after a year on Arveon.

"We moved 12 services off a hyperscaler. The migration took a long weekend. Egress savings paid for the year inside six weeks."
Elin Lindqvist
VP Infrastructure · Tidewave
"The console is the first one in years where I haven't sworn at the console. Plain CLI. Plain Terraform. Plain bills."
Joonas Mäkinen
CTO · Numerix.dev
"Sub-millisecond inter-AZ on day one. Real engineers answering real PagerDuty pages. Refreshing."
Nadia Kowalski
Principal SRE · Cardinal Data
Pricing

Pay by the second.
Read the price before the bill.

Hatch

a2

Starter VPS. Personal projects, side experiments.

€16/mo
€0.022 / hour · €0.0000061 / second
  • 2 dedicated vCPU (AMD EPYC)
  • 4 GB ECC RAM
  • 80 GB NVMe SSD
  • 4 TB outbound transfer
  • Anycast IP · IPv6 /64
Start with Hatch
Anvil

a24

Heavy workloads, databases, render fleets.

€420/mo
€0.583 / hour
  • 24 dedicated vCPU
  • 64 GB ECC RAM
  • 1.2 TB NVMe SSD
  • Unlimited outbound transfer
  • 99.99% SLA · priority support
Start with Anvil
No long forms · No sales calls

Deploy in twelve seconds.
Stop paying for egress.

€10 in trial credit on signup. No credit card. No "talk to sales" gating. Production workloads from day one.

or: Read the docs See pricing