The Swedish cloud that tells it like it is.
Our own stack. Green Nordic power. An API and a console you actually like on day one. Kepler sits exactly where it hurts — in the seam between your app and the infrastructure — and makes the seam disappear.
A reply from a Swedish engineer within 4 hours, on weekdays.
Hyperscalers give you 200 Lego bricks. We give you a Planet.
A Planet is the power of OpenStack wrapped in the Kepler layer — compute, networking and storage fused with the experience: console, API, identity, quotas, isolation and billing. An object you actually want to use.
OpenStack
- Nova — compute
- Neutron — networking & SDN
- Cinder — block storage
- Glance — OS images
The Kepler layer
- Console & REST API
- Identity, teams & RBAC
- Per-tenant isolation (OpenBao)
- Quotas, audit log, billing
Your isolated part of Kepler in a region — your own tenant project, your own network, your own storage and your own credentials. "Spin up a Planet and you have your own cloud in Falkenberg."
App DevOps on one side. Infrastructure on the other. We erase the seam.
App DevOps
Deploys, environments, CI/CD, scaling. What you actually care about.
The gap
200-service consoles, billing surprises, certifications you need a consultant to read.
Infrastructure
Compute, networking, storage, isolation. The things that simply have to work.
Small, sharp and honest. This runs in production today.
Everything below exists in the console and the API right now. No promises, no asterisks. What sits on the roadmap we say on the roadmap page — not here.
Compute Now
Create, delete, resize, restart VMs. Boot-from-volume, curated OS images, noVNC console, SSH keys.
ComputeNetworking Now
Provider networks with direct public IP, VPC, routers, NAT gateway, floating IPs and security groups.
NetworkingBlock storage Now
Block volumes, volume and server snapshots. Attach, resize, snapshot, restore — via console or API.
StorageTeams & RBAC Now
Four built-in roles plus custom. API keys, REST API and a console that mirror each other exactly.
PlatformAPI + Terraform Now
Idempotent REST API, async audit log, quota enforcement. Everything the console does, you can script.
API & TerraformPer-tenant isolation Now
A dedicated OpenStack project per tenant, with credentials managed by OpenBao. A checkable security proof.
Security modelFrom terraform apply to a running instance — in 47 seconds.
resource “kepler_instance” “api” {
region = “swe1” # Falkenberg
flavor = “gp1.large”
image = “ubuntu-24.04”
network = kepler_network.app.id
ssh_keys = [kepler_ssh_key.deploy.id]
}
$ terraform apply
kepler_instance.api: Creation complete (47s)
→ floating_ip = “194.x.x.18”
A lightweight illustration of the flow — the real console and our official Terraform provider are the same control surface we run ourselves. See the full stack →
Take Kepler with you everywhere.
Console, CLI and API — the same control surface. Start, resize and snapshot your Planets from the couch, the pipeline or the terminal.
Every resource, exactly where you left it.
The console mirrors the API exactly — instances, volumes, networks and snapshots in a view that doesn't require a certification to read.
Our own hardware in a 100% renewable Nordic data center — your data stays in Sweden.
We run our own servers and our own cloud stack inside a 100% renewable Nordic data center in Falkenberg. The stack and the servers are ours; the renewable power and waste-heat recovery belong to the facility.
That means a concrete answer to the data residency question: your data lives in Sweden, under Swedish and EU law. Per-tenant isolation with OpenBao-managed credentials is checkable proof — not a marketing phrase.
100% renewable Nordic power
Waste heat is recycled into district heating — via the facility in Falkenberg.
Data in Sweden
Under Swedish and EU law. GDPR-compliant by design.
ISO 27001 + SOC 2 starting Aug 2026
We say "program in progress" — never "certified" before the certificate exists.
Small today. Growing fast — and we're honest about what doesn't exist yet.
Kepler is building to become the obvious platform for SaaS companies and startups to launch their apps on. This is on the roadmap — clearly labeled, never sold as the present.
GPU compute Now
6× NVIDIA L40S — single-tenant, booked via a call. H100-class: roadmap.
Apps / App Engine Roadmap
Managed WordPress / Laravel / Node / static.
K8s · object storage · LB Roadmap
Managed Kubernetes and load balancers. Kepler-hosted object storage — today via a European partner.
LLM & Claude Code Vision
Kepler's own direction toward a launch platform for AI teams. No third-party partner.
Spin up your first Planet today.
Real prices, an honest roadmap, and a Swedish engineer who actually replies. Need an RFP pack for procurement? We have one ready.