HCP Terraform against self-hosted Terrakube for teams that want the same organization, workspace and private registry model: remote backend and cloud block support, VCS integrations, policy checks with OPA, SSO through Dex, and running OpenTofu instead of BSL Terraform.
HCP Terraform against Atlantis for teams that review infrastructure changes in pull requests: plan output as comments and directory locking versus a hosted run UI, what Atlantis does not provide (a registry, workspace UI or managed state), and the state backend Atlantis requires.
Moving workspaces from HCP Terraform to Terrakube: downloading current state through the state versions API, pointing the cloud block or remote backend at the Terrakube API, re-creating variables, VCS links and teams, and switching the CLI to OpenTofu where wanted.
Moving from HCP Terraform runs to pull request automation with Atlantis: choosing a state backend such as S3 or GCS (Atlantis cannot use local state), moving each workspace's state, writing atlantis.yaml for modules and workspaces, and running OpenTofu through a custom workflow.
Cost of ownership by component: HCP Terraform's per-managed-resource pricing, what the 500-resource Free tier covers after the legacy plan ended, and Terraform Enterprise quotes, against Terrakube or Atlantis infrastructure, state storage, setup and the platform team that runs it.
Getting state, variables, workspace settings and registry modules out of HCP Terraform: the state versions API and its raw and JSON download URLs, which objects have no export and must be re-created, and checking every workspace before cancelling.
For teams already running Terraform Enterprise on their own Docker, Kubernetes or OpenShift: Terrakube as the self-hosted replacement, compared on remote runs, the cloud block, private registry, identity through Dex, policy and cost extensions, licensing and support.
Moving workspaces from self-managed Terraform Enterprise to Terrakube: exporting current state through the state versions API, recreating organizations, workspaces, variables and VCS connections, switching the cloud block or remote backend, and moving to OpenTofu at the same time or later.
Replacing Sentinel and HCP Terraform's policy checks with Open Policy Agent: Terrakube's OPA extension in runs, or Atlantis policy checks that run conftest on the plan and need policy-owner approval to override, and how Sentinel policies translate to Rego.
Where private modules and providers go when HCP Terraform does: Terrakube's built-in private registry, the public OpenTofu Registry for community providers, and how module sources and version constraints change in your configurations.
HCP Europe requires a new account and organization, cannot take migrated organizations and leaves out teams and some integrations. This page compares that path with running OpenTofu and Terrakube in an EU region or data center you control.
The legacy Free plan ended on March 31, 2026 and organizations moved to the enhanced Free tier of 500 managed resources. This page explains how managed resources are counted, when growth pushes you to paid tiers, and when self-hosted OpenTofu with Terrakube or Atlantis is the cheaper path.