Terraform pricing, alternatives, and review

The short answer

Buy Terraform if you already write Terraform and want shared state, access control and a private module registry run for you, and you can count the resources under management closely enough to predict a bill that scales with your infrastructure.

Skip it if a handful of engineers run the CLI against remote state you already host, or if your estate is large and the per-resource meter prices it badly. The CLI and the hosted platform are separate purchases, and only the second one is metered.

Where Terraform fits

Terraform is infrastructure as code: you declare resources in configuration files, providers translate that into API calls against the platform in question, and a state file tracks what exists. HashiCorp's documentation calls state the source of truth for your environment and describes providers as what enables Terraform to work with virtually any platform or service with an accessible API, with thousands of them in the public registry. Modules are the reusable units people build to stop writing the same thing twice.

The split that decides the bill is CLI against platform. The CLI is what runs on your machine or in a pipeline. HCP Terraform is the hosted service HashiCorp describes as running Terraform in a managed environment, adding secure access to shared state and secrets, role-based access controls and a private registry for modules and providers. The documentation page we read does not state the licence for the CLI, so this page does not assert one either way.

For an IT operations team, the honest fit is narrow but real. If your infrastructure is provisioned by hand through a console, Terraform's value is that a change becomes a reviewable diff and a destroy becomes an explicit plan. That is the same argument as for configuration management tools such as Puppet Enterprise or Chef Infra, with one important difference: Terraform provisions and tracks resources, where those two drive the ongoing state of machines you already have.

Where it stops being an IT ops tool is day-two operations on servers. Terraform will create a virtual machine; it will not keep its packages patched, manage its configuration drift or tell you it is unhealthy. Teams that adopt it for provisioning still need configuration management, monitoring and endpoint tooling, and the DevOps automation tools category is where the rest of that chain sits.

The other fit question is who holds the state. Shared state is the single most valuable thing HCP Terraform sells, because two engineers running the CLI against the same infrastructure with local state is how estates get corrupted. You can solve that yourself with a remote backend, and many teams do. What you are buying from the hosted product is not the capability but the operation of it, plus the access control and registry around it.

Terraform pricing

HashiCorp prices HCP Terraform per resource under management, and the page says nothing about per-user pricing. Essentials starts at $0.10 per month per resource, Standard at $0.47 and Premium at $0.99, each also quoted as an hourly rate. Terraform Enterprise, the self-managed edition, and HCP Terraform Europe both read Custom Pricing. Checked on October 16, 2026.

What the HashiCorp pricing page publishes for Terraform

HCP Terraform Essentials

Starting at $0.10 per month per resource, quoted as $0.00013 per hour. Includes a $500 HCP trial credit with no published duration.

HCP Terraform Standard

Starting at $0.47 per month per resource, quoted as $0.00064 per hour. Includes a $500 HCP trial credit.

HCP Terraform Premium

Starting at $0.99 per month per resource, quoted as $0.00135 per hour. Includes a $500 HCP trial credit.

Terraform Enterprise, self-managed

Custom Pricing. No figure is published, and the route is to sales.

HCP Terraform Europe

Custom Pricing. Also quote-only, which matters if data residency is a requirement.

Free tier

The page does not state whether one exists. It carries an FAQ question asking exactly that, and the answer is not in the page content, so this page makes no claim either way.

How the bill behaves as managed resources grow

This meter is worth doing the arithmetic on, because almost nobody can estimate it in advance. On Standard at $0.47 per resource per month, 1,000 managed resources is $470 a month, or $5,640 a year. The same 1,000 resources on Essentials at $0.10 is $100 a month, or $1,200 a year, and on Premium at $0.99 is $990 a month, or $11,880 a year. Those are products of the published starting rates and a round resource count, not quotes.

The reason it surprises people is that a resource is not a server. A single module that stands up one application can declare a virtual machine, a disk, a network interface, a security group, a DNS record, several IAM objects and a load balancer rule, and each one is a resource under management. The ratio of resources to things a human would count is commonly five or ten to one, which means the cost driver is how your configuration is written as much as what infrastructure you own.

Two consequences follow. First, every rate on the page reads starting at, so these are floors rather than quotes, and the real number comes from a conversation. Second, the meter rewards tidiness: unused resources left in state cost money, and splitting one workspace into many does not reduce the count. Before you commit, run a state list against your existing configuration and count. That number, multiplied by the tier rate, is your budget.

A different meter in the same category

For contrast on the same day, GitLab published $29 per user per month for Premium and GitHub published $21 per user per month for Enterprise, both per seat. Those are those vendors' figures. The point is not that Terraform is cheaper or dearer but that it cannot be compared on a seat basis at all: a five-person platform team managing a large estate pays more here than a fifty-person team managing a small one, which is the opposite of how the rest of this category behaves.

Count your resources before you talk to anyone, and treat the published rates as floors because every one of them says starting at. Ask explicitly whether a free tier exists, since the page poses that question itself and does not answer it, and ask what the $500 trial credit's duration is. If the business case is about safer changes to infrastructure, the IT downtime cost calculator and the incident postmortem template are the two things that make that argument concrete.

Three caveats, recorded when the page was checked on October 16, 2026. Every tier is phrased as starting at, so the published rates are floors rather than quotes. The page publishes no per-user pricing of any kind, which is unusual enough in this category to be worth repeating. And the page does not state whether a free tier exists: it carries an FAQ question asking that, and the answer is not in the page content, so nothing is asserted here either way. Terraform Enterprise and HCP Terraform Europe are Custom Pricing.

Checked October 16, 2026 against the HashiCorp pricing page.

What Terraform gives up

You give up a predictable unit. Per resource under management means the bill tracks the shape of your configuration, not the size of your team, and the count moves every time somebody adds a module. Teams used to seat licensing should expect their first invoice to be a learning experience rather than a confirmation.

You give up a published price for the self-managed edition. Terraform Enterprise reads Custom Pricing, which is the option regulated organisations most often need, so the people with the strictest requirements get the least information from the page. HCP Terraform Europe is quote-only too, which makes data residency a sales conversation rather than a tier choice.

You give up certainty about the free tier. HashiCorp's own page asks whether a free version exists and does not answer it where we could read it. That is a strange gap, and it means anyone planning to start small has to confirm the entry point with the vendor rather than from the page.

You give up day-two management. Terraform provisions and tracks; it does not patch, monitor or remediate. The gap between create the server and keep the server correct is filled by configuration management and monitoring tools, and pretending otherwise is how estates end up with drift nobody notices.

And you give up some operational safety if you skip the hosted product. Shared state and access control are the real value of HCP Terraform, and the alternative is building a remote backend, locking, secret handling and a module registry yourself. That is entirely doable and it is work, which is the same trade that Jenkins under its MIT licence presents in CI.

How Terraform compares

Terraform meters resources where almost everything else in this category meters people or minutes, so the table below is mostly a comparison of units. Prices are only figures verified on that vendor's own page on October 16, 2026; an unchecked sibling says so in its row.

How to test Terraform in two weeks

Each paid tier includes a $500 HCP trial credit with no published duration, so the first job is to ask how long it lasts. The second is to find out what you will be billed on, because the meter is resources and most teams have never counted theirs.

1. Count first. Run a state listing against your existing configuration, or write the configuration for one real application and count the resources it declares. Multiply by $0.10, $0.47 and $0.99 to see the three tiers in your own numbers.
2. Ask HashiCorp two questions in writing: how long the $500 trial credit lasts, and whether a free tier exists, since the pricing page poses that question itself without answering it.
3. Move one workspace's state into HCP Terraform and have two engineers work against it at once. Shared state and locking are the core of what you are buying, so test the conflict case deliberately rather than the happy path.
4. Publish one internal module to the private registry and consume it from a second workspace. If your team will not use the registry, a chunk of the tier difference is wasted.
5. Run a plan and apply through a pipeline rather than from a laptop, with the approval step your change process requires. The audit trail is the part an operations team is usually buying.
6. Before deciding, tidy the state: destroy what you no longer use and re-count. On a per-resource meter, housekeeping is a cost control, and knowing how much it is worth is part of the evaluation.

Frequently asked questions

What is Terraform used for?

Infrastructure as code. You declare resources in configuration files, providers translate that into API calls against the platform in question, and a state file tracks what exists. HashiCorp calls state the source of truth for your environment, which is what turns an infrastructure change into a reviewable diff rather than a console click.

Is Terraform free or paid?

Both exist and the pricing page will not settle it. HCP Terraform, the hosted service, is paid from $0.10 per month per resource. Whether a free tier exists is not stated: the page carries an FAQ question asking exactly that and the answer is not present in the page, so this page asserts nothing either way.

How much does Terraform cost?

HCP Terraform meters resources under management: Essentials starting at $0.10 per month per resource, Standard at $0.47 and Premium at $0.99, each quoted with an hourly equivalent. Terraform Enterprise, the self-managed edition, and HCP Terraform Europe both read Custom Pricing. Verified October 16, 2026.

Why is Terraform priced per resource instead of per user?

Because the hosted product bills for what it manages rather than who logs in, and the page publishes no per-user figure of any kind. The practical effect is that cost tracks the size of your infrastructure: a thousand managed resources on Standard is $470 a month regardless of how many engineers touch it.

What counts as a managed resource?

The pricing page meters resources under management and does not define the term on the page itself, so nothing precise is asserted here. Since the whole bill turns on that count, ask HashiCorp how resources are counted for your configurations, including whether data sources and ephemeral resources are included, before committing.

Is Terraform for AWS only?

No. Providers are what let it work with a platform, and HashiCorp describes them as enabling Terraform to work with virtually any platform or service with an accessible API, with thousands of providers in the public registry. AWS is one of many, and the same configuration style covers all of them.

Is Terraform still relevant in 2026?

On the evidence of the vendor's own page, yes: HashiCorp was actively selling and pricing HCP Terraform on October 16, 2026, with published per-resource rates across three tiers and two quote-only editions. ITCommons has verified no end-of-life notice and no replacement product, and will not repeat a migration trend it has not measured.

What is the difference between the Terraform CLI and HCP Terraform?

The CLI runs on your machine or in a pipeline. HCP Terraform is the hosted service HashiCorp describes as running Terraform in a managed environment, adding secure access to shared state and secrets, role-based access control and a private registry for modules and providers. Only the hosted service is metered.

Is Terraform Enterprise priced differently?

It is not priced publicly at all. Terraform Enterprise, the self-managed edition, reads Custom Pricing and routes to sales, as does HCP Terraform Europe, which matters if data residency is a requirement. So the published per-resource rates describe the hosted product only.

Does Terraform replace configuration management tools?

No, it sits beside them. Terraform provisions resources and tracks them in state. Tools such as Puppet Enterprise and Chef Infra drive the ongoing state of machines you already have. Terraform will create a virtual machine; it will not keep its packages patched.