Skip to main content
Control PlaneControl Plane
Control Plane is a cloud virtualization platform. It combines regions of AWS, GCP, Azure, OCI, and other providers with bare metal, on-prem VMs, and Kubernetes clusters you already run into one or more Global Virtual Clouds (GVCs). Your workloads, whether containers, VMs, or AI agents, run in every location their GVC spans and get the same networking, identity, secret management, observability, and access control in each one. You operate all of it through the Console, the CLI, the REST API, Terraform, Pulumi, and an MCP server.

Start here

Deploy your first workload

Create a Global Virtual Cloud (GVC), run a sample application as a workload, and open it on a global TLS endpoint.

Understand the concepts

Read these pages first to learn what each building block is and how they fit together.

Follow a how-to guide

Configure a domain, create a secret, push an image, set up a CI/CD pipeline, and more.

Point an AI agent at it

Which path your tool takes: the AI Plugin, or the MCP server.

Compliance

PCI DSS Level 1 and SOC 2 Type II: the attestations, the audit scope, and the responsibility matrix.

Build with AI

Sandboxes

Isolated environments for AI coding agents, reachable from a browser IDE, VS Code, Cursor, or SSH.

AI Plugin

Skills, agents, and guardrails that teach Claude Code, Codex, and Antigravity CLI to drive Control Plane.

MCP Server

Connect Cursor, Claude Desktop, VS Code, or any other MCP client to your org.

Run a model

Ollama, LiteLLM, Qdrant, and Langfuse install from the Template Catalog as versioned releases.

Build and deploy

Create a GVC

Pick the locations your workloads run in. Workloads, identities, and volume sets live inside it.

Create a Workload

Run a container as a serverless, standard, stateful, cron, or VM workload.

Push an image

Build into your org’s free private registry, or push an image you already have.

Create a Secret

Credentials, keys, and certificates, plus the identity and policy chain a workload reads them through.

Configure a Domain

Route your own domain to a workload. TLS certificates are issued and renewed for you.

Install from the Template Catalog

PostgreSQL, Redis, Kafka, and more, installed into a GVC as versioned releases.

Run a full virtual machine

A VM workload brings its own kernel, init system, and guest OS, including Windows. Lift an existing disk image across without re-platforming it, and it still gets the same encrypted internal networking, the same credential-free access to cloud services, and the same firewall rules as your containers.

Interfaces

Console

The web UI. Explore resources, make one-off changes, and read logs and metrics in the browser.

CLI

cpln scripts the same operations, and adds exec, port-forward, and cpln apply manifests.

REST API

Every operation over HTTP at api.cpln.io, generated from an OpenAPI spec.

Terraform

The official cpln provider, managing every resource kind in HCL.

Pulumi

The same resources in TypeScript, Python, Go, or C#.

MCP Server

The same operations over MCP at mcp.cpln.io, for an AI agent to call directly.

Kubernetes Operator

Already running a cluster? Manage Control Plane resources as CRDs, so kubectl and ArgoCD drive them alongside your Kubernetes manifests.

Infrastructure

Managed Kubernetes

Control Plane provisions and operates the clusters, on your cloud accounts or your own servers.

CPLN BYOK

Bring your own Kubernetes: register a cluster you already run and it becomes a location workloads roll out to.

Wormhole agent

Reach TCP and UDP endpoints inside a VPC, an on-prem data center, or a developer’s laptop.

Native networking

AWS PrivateLink and GCP Private Service Connect, for services that never leave the VPC.

Observe and automate

Logs

Workload, access, and cron logs, queryable with LogQL from the Console, CLI, API, Grafana, or an AI agent.

Metrics

What every org, workload, replica, and volume emits before you instrument anything.

Audit Trail

Every platform action, plus your application’s own events written to the same trail.

External logging

Forward org logs to S3, CloudWatch, Datadog, Coralogix, Loki, and other providers.

CI/CD

Working pipelines for GitHub Actions, GitLab CI, Bitbucket Pipelines, and CircleCI.

Tracing

Follow one request across every workload it passes through, at a sampling rate you set.

Reference documentation

Workload

Containers, autoscaling, firewall, volumes, load balancing, and termination.

GVC

Locations, pull secrets, the shared identity, and GVC-wide environment.

Identity

Credential-free cloud access, and the network resources a workload may reach.

Policy

Bindings, permissions, and the principals a policy grants them to.

Domain

Routing modes, routes, ports, TLS settings, and CORS.

All resources

Every kind, including secrets, volume sets, images, locations, cloud accounts, and quotas.
Need help? Read the FAQ, watch the Product Demos, or catch up on the Blog.