Scalr review

Scalr is a hosted SaaS control plane for running Terraform and OpenTofu workflows. It sits between your source repository and the infrastructure providers Terraform manages: connect workspaces, configure variables and state, then run plans and applies through Scalr. The service is positioned as a drop-in Terraform Cloud replacement, and its migration guidance says to change the hostname and organization in the cloud block while continuing to use Terraform CLI commands.

Scalr adds workflow and governance features around those runs: PR-native plans and applies, policy checks with Open Policy Agent, Checkov or both, drift detection, and account-wide reporting. Workspaces can run on Scalr or on self-hosted agents in your own infrastructure. The hosted service is not a Terraform library or a self-hosted Scalr server. For Enterprise, the vendor says state can be stored in your own bucket; otherwise, Scalr can store it. This makes Scalr an orchestration and management layer, not a replacement for Terraform configurations or cloud providers.

The documented integration path is to point Terraform code at Scalr by changing the hostname and organization in the cloud block. The homepage says existing Terraform CLI commands work as before, and Terraform 1.6+ can move directly to OpenTofu. For Git-based workflows, Scalr offers apply-before-merge, pull-request run summaries, /scalr plan and /scalr apply slash commands, and branch-aware state protection.

The vendor also provides an MCP server over HTTP for Claude, Cursor or VS Code. Its example connects Claude Code with claude mcp add --transport http and an account-specific Scalr endpoint; authentication happens from a session. MCP tools use the account user's permissions, and most are read-only. The homepage does not name SDKs or programming languages for a general-purpose API, so confirm SDK support and API coverage in Scalr's docs before building custom integrations. It points developers to documentation for flexible workflows and mentions custom workflow hooks, but does not describe webhook support on the homepage.

The Free plan costs $0 and includes up to 50 runs per month, all core features, and no charge for users, resources, concurrency or self-hosted agents. That makes it suitable for evaluation and small workloads, but teams exceeding 50 monthly runs need to account for paid usage.

Business costs $0.99 per run, with volume discounts available. The pricing page lists annual plans, but not their prices. There are no listed per-user, managed-resource, concurrency or self-hosted-agent fees, so usage growth is driven mainly by billable runs. Runs beyond prepaid plans cost $0.99 each and are billed at the end of the subscription period.

Enterprise has custom pricing and starts at 20,000 runs per year. It adds all core and enterprise features, a Technical Account Manager, a shared Slack support channel, and higher dedicated CPU and memory resources. Ask Scalr for a quote based on expected run volume and confirm how prepaid runs and excess usage apply to your contract.

Scalr is sold as a SaaS platform. Its self-hosted agents run workflows on your infrastructure; that is not the same as self-hosting the Scalr control plane. The homepage does not identify an open-source licence or describe a self-hosted server, so teams that require either should ask Scalr directly rather than assume the product can be deployed independently.

The vendor's migration guide describes moving Terraform Cloud projects, workspaces, variables and state to Scalr in batches. For Enterprise, state can be stored in your own bucket; the homepage does not describe a general export format or a complete exit procedure. Terraform configurations remain the basis of the workflow, but workspace settings, state handling, policies and PR integrations may need work to move elsewhere. Before migrating production, confirm state export, API-token portability, and how to retrieve workspace and run data if you leave.

Scalr's billing model avoids per-resource and concurrency charges, and the vendor says self-hosted agents are included at no extra cost. That can simplify cost estimates, but run volume still matters: Business is billed per run, and runs beyond prepaid plans cost $0.99 each. Enterprise starts at 20,000 runs a year and includes higher dedicated CPU and memory resources; ask how those resources and run limits map to your workloads.

The platform includes drift detection, policy checks, RBAC, SAML, OIDC and secrets management. Audit logging is listed for Enterprise. Scalr also reports on run duration, queue and approval wait, provider and module versions, stale workspaces, resources and API tokens. These features help teams identify operational issues, but the homepage does not publish service-level or rate limits. Before depending on Scalr, confirm API limits, availability commitments, backup and recovery procedures, data-export options, and the support response terms for your plan. For private-network workloads, validate agent connectivity and capacity with the vendor.