Scalr vs Terraform Cloud Pricing: Per-Run vs Per-Resource (RUM)

As of September 2026, Scalr bills per-run and HCP Terraform (formerly Terraform Cloud) bills per-resource, on resources under management (RUM), a meter HashiCorp adopted in June 2023 in place of per-user, per-apply pricing. Scalr is free to 50 runs a month, then $0.99 a run with volume discounts on prepaid packages. HCP Terraform's tiers start at $0.10, $0.47, or $0.99 per resource per month, after a 500-resource free allowance. Which is cheaper comes down to one question: does your run count or your resource count grow faster? Scalr's bill tracks activity, so it holds steady as your estate grows. Terraform Cloud's rises with the resource count whether or not anything changed.

What are the plans on each side, and what does each one meter?

Scalr

Scalr has three plans: Free, Business, and Enterprise.

Free covers 50 runs a month, resetting on the first of the month, with community support. Concurrency on Scalr's managed runner pool is a quota, not a paid feature: as of September 2026 Scalr's pricing FAQ says new accounts get five concurrent runs by default, raised free of charge through a support ticket with a valid use case. You can also deploy as many self-hosted agents as you like on any plan, and each one adds five concurrent runs (a ratio chosen to keep run execution reliable).

Business is the plan for teams that will pass 50 runs a month. Same core features as Free, standard support, and a trial if you want one. Every object in Scalr has a quota to prevent abuse, and on paid plans any quota, concurrency included, is raised free after a look at the use case.

Enterprise adds the controls larger shops ask for: data location, your own encryption keys, identity, audit trail, and enterprise contracts. Dedicated run capacity isn't an Enterprise-only feature; self-hosted agents give you that on any plan at no extra cost.

The only question you need to answer on Scalr is whether you'll pass 50 runs a month.

Terraform Cloud

HashiCorp lists three public tiers for HCP Terraform (Essentials, Standard, Premium) plus Terraform Enterprise, the self-hosted product. The legacy self-serve free plan reached end of life on March 31, 2026; an enhanced free tier continues at up to 500 managed resources and one concurrent run. Every paid tier is priced per resource under management.

What separates the tiers is feature gating and concurrency. These sit behind the higher tiers, and all of them are on Scalr's free plan: team management, no-code provisioning, cross-organization module sharing, drift detection, ephemeral workspaces, and run metrics.

Essentials applies the RUM model to a basic feature set for small teams: OPA enforcement is limited to a single policy set with one mandatory policy, and run tasks to one integration. Standard adds team management, keeps the same policy and run-task limits, and caps you at three concurrent runs per IBM's limits page (Free is one; no integers are published for Essentials or Premium). Premium unlocks the rest: no-code provisioning, audit logs, drift detection, ephemeral workspaces, unlimited OPA policy sets and enforcement levels, unlimited run-task integrations. Premium is publicly listed, not quote-only.

Terraform Enterprise runs inside your own network and has no published price. From what we can gather it's priced per workspace and doesn't follow the RUM structure, but you'll need to talk to sales to find out.

So on HCP Terraform you're answering three questions at once: how many resources you have, which features you need (audit trails and drift detection force you to Standard or above, no-code provisioning isn't in Essentials), and whether the concurrency cap on that tier will queue your team. Any one of them can push you into a higher per-resource rate than your estate size alone suggests.

What do you commit to?

Scalr offers monthly or annual terms. Monthly, you prepay a number of runs you expect to use, and can move up or down at any time; anything over the prepaid amount is billed at the $0.99 flex rate until the next cycle. Annual, you buy runs in bulk with a year to use them and a volume discount on the whole purchase, and if you run out early you renew early or pay flex until renewal. Either way the only thing you commit to is a number of runs, and every discount is in the pricing calculator.

HCP Terraform's pricing page shows pay-as-you-go across the public tiers at the per-resource rates above (first 500 resources free) and annual commitments on the paid tiers. Terraform Enterprise is quote-based.

Which runs are billed on Scalr's per-run meter?

On Scalr, a run is the only thing you pay for. There's no charge for support, new features, or raising quotas. And not every run counts: drift-detection runs, runs stopped by a pre-plan policy or Checkov, runs that fail during init or cost estimation, and runs that fail on an internal Scalr error are excluded. The full list, keyed by error code, is in the billing FAQ.

On HCP Terraform the headline price is per resource per month but the meter runs hourly, with partial hours rounded up, so a resource that exists for a week bills for a week.

What does support cost?

Scalr has two support tiers, free and paid. Free is ticket-based and best-effort. Paid is 24x7x365 with response times by ticket priority, plus the support portal, a Slack channel, and screen sharing as needed, at no extra cost on the paid plans.

As of July 2026 the HCP Terraform pricing page doesn't publish per-tier support details. Terraform Enterprise is listed with premium support included; the rest is worked out with sales. If responsiveness matters to you, get it in writing during the quote.

How do you estimate your Scalr bill?

The tfc-runs script pulls every run ever executed in your Terraform Cloud account. Divide by the account's age to get a monthly figure (the API has no date filter, so you'll estimate), then put that number into the pricing calculator.

What does a drop-in migration from Terraform Cloud involve?

Scalr runs the same remote-operations-backend model as Terraform Cloud (the cloud/remote backend, the same CLI and API), which is what makes the move a drop-in rather than a re-platforming: workspaces, variables, and state move over, and the terraform commands and pipelines your team already uses keep working. Sensitive variable values are the exception, since the Terraform Cloud API won't export them and they're re-entered by hand. Three routes: the migration module automates moving objects from Terraform Cloud; you can pull state with the Terraform CLI and push it into a Scalr workspace by hand; or, for larger organizations, build the Scalr shell first (environments, teams) and let individual teams migrate by a set date. The self-hosted options post has more on the mechanics, and the pillar on selecting a Terraform Cloud alternative has the full comparison set. Scalr offers free migration periods to some teams leaving Terraform Cloud. Ask when you talk to us.

How do the other TACOs meter: workers, applies, or self-hosting?

Scalr and HCP Terraform are TACOs (Terraform automation and collaboration software). The Spacelift figures below were re-verified on September 5, 2026, and env0's in August 2026; Atlantis is open source.

Atlantis

Open source, so the only cost is running and maintaining the installation. It's aimed at GitOps users and integrates with all the major VCS providers.

env0

env0 is a CI/CD platform for Terraform, OpenTofu, Pulumi, and CloudFormation. As of August 2026 its pricing page lists a free tier and two paid tiers, Cloud Navigator and Cloud Pilot. The free tier is $0 in perpetuity: 250 runs a month across up to 30 active environments, unlimited users, agents, and concurrency, with what env0 describes as the full Cloud Navigator feature set including drift detection and OIDC SSO, and community support only. env0 counts a run as one successful apply or one drift detection, so plans and failures aren't metered. This replaced a period earlier in 2026 with a trial and no perpetual free plan, and the $1,500/month Cloud Compass tier that once headlined the page is gone (Cloud Compass is now a feature group).

Both paid tiers bill per successful apply or per environment and have no published list price. They buy higher ceilings (100 active environments on Cloud Navigator, unlimited on Cloud Pilot), direct support, and the capabilities reserved for Cloud Pilot: Cloud Analyst, advanced drift analysis, cost management, and self-hosted agents.

Spacelift

Spacelift is a CI/CD platform for Terraform, OpenTofu, Ansible, Pulumi, and CloudFormation. As of September 2026 it lists four paid plans (Starter+, Business, Enterprise, Enterprise+) with different features and support levels, and beyond features the pricing unit is private workers, which execute your deployments. A worker processes one run at a time, so Spacelift is concurrency-metered: you're buying parallel run slots, and the bill scales with how many you hold. Users are unlimited on every paid plan, so seats aren't the meter. Moving there from Terraform Cloud is a re-platforming, not a drop-in: Spacelift's docs say not to specify a Terraform backend at all, and its migration tool turns Terraform Cloud workspaces into stacks. If you run several IaC tools, that breadth is the reason to pick it.

That model has a structural problem compared with per-run pricing. A fixed slot count is always the wrong number: too few and engineers queue during releases, too many and you rent idle capacity. It fails hardest during an incident, when parallel fixes queue behind the cap and getting another slot means procurement under pressure. And the capacity planning lands on you, forecasting peaks and watching utilisation. Per-run pricing has none of that: you pay for runs that executed, and there's no slot to mis-provision.

The only public price is the entry paid tier, Starter+, at $20,000/year (unlimited users, 2 public workers, 1 private worker, with a second private worker available). The free plan covers 2 users and 1 public worker. Business, Enterprise (5 to 30 private workers), and Enterprise+ (12 to 30) are quote-only, and SSO / SAML 2.0 requires the Enterprise plan. Drift detection is a Starter+-and-above feature that, per Spacelift's docs, only works on private workers; on Scalr, scheduled drift-detection runs aren't billed on any plan. Older "$399/month Starter" figures on third-party sites are stale.

For the full teardown of HCP Terraform's own pricing, see Terraform Cloud pricing revealed.