Why you should switch to Stategraph today
Usual Terraform state management hands you a global lock, a full-state refresh, and a JSON blob with dependencies buried inside attribute strings. Stategraph replaces the file with a graph and the global lock with resource-level locking.
Run plans only on what changed
Stategraph computes the minimal change cone for every plan. A modification to three resources in a 2,847-resource state evaluates those three, not all 2,847. Plans that used to drag through full state refreshes complete in seconds.
Apply in parallel without lock contention
Two teams touching independent resources in the same state don't queue behind a global lock. Stategraph only locks the subgraph each operation needs. Zero contention on disjoint changes. Conflicts surface cleanly at commit.
Manage resources across multiple states as one unit
Multi-state atomic transactions mean a change in one state that affects downstream states gets planned and applied together in one operation. No manual orchestration, no Terragrunt wrappers, no sequential state-by-state apply loops.
Query your infrastructure like a database
State lives in PostgreSQL, so you can run SQL directly against it. Blast radius, inventory, compliance checks, and cost context are all derived from the same structured model, not parsed from flat files after the fact.
Keep your existing Terraform configuration files
Stategraph is a drop-in replacement. Your HCL, modules, providers, Terraform commands, and existing state files all stay. Import with one command, replace terraform with stategraph, and you are running.
Why Stategraph is so much faster
Infrastructure as a Database is not a different implementation of Terraform's planning algorithm. It is a different algorithm. Instead of loading, locking, and refreshing an entire state file on every operation, Stategraph operates on the affected subgraph: the minimal set of resources your change actually touches.
| Scenario | Terraform (traditional) | Stategraph |
|---|---|---|
| No-op plan on 800 MB state | 3+ hours | 97 seconds |
| Small change in large state | Minutes (full refresh) | 2.3 seconds |
| Two teams, same state file | Queue behind a global lock | Parallel, zero contention |
Measured on real production infrastructure. 800 MB state file, partner production environment. Single benchmark; results vary with state shape.
Subgraph execution replaces full-state refreshes
When you run a plan, Stategraph computes the minimal change cone: only the resources your change actually touches. Terraform has no concept of this. It refreshes the entire state on every operation. The state is a flat JSON file and dependencies are buried inside attribute strings.
Resource-level locking replaces the global lock
Terraform acquires a global lock when a plan or apply runs. Every other operation on that state file waits. Stategraph locks the resources a given operation touches. Two engineers can work on independent parts of the same state in parallel. When changes overlap, the second transaction is rejected at commit, meaning no corrupted state.
PostgreSQL replaces the flat file
Infrastructure state stored as rows in a relational database rather than a flat JSON blob means the query is proportional to the subgraph, not the file. For small changes in large infrastructure, this is why the numbers look so different. It is a different algorithm, not just a faster implementation of the same one.
Detect drift without extra tooling
Configuration drift detection is built into the Stategraph orchestration layer. Scheduled refreshes flag when reality has diverged from your desired state, and the results land in the same system of record as everything else.
See the difference
Watch how Stategraph parallelizes your infrastructure operations.
Use Stategraph Cloud as your Terraform SaaS
The Free tier is shared-tenant SaaS with no time limits. Import your state, explore the graph, and see blast radius immediately. Paid tiers are single-tenant: your own isolated Postgres instance, never sharing a database with another customer.
Stategraph Cloud pricing
All features included on every plan. Start free, move to Orchestration Pro to remove every limit, and add Infrastructure as a Database from $25,000 a year.
- All features included
- Plan and apply from pull requests
- 3 users
- 50 runs per month
- 1 concurrent worker
- Community support
- Everything in Free, no limits
- Unlimited users
- Unlimited runs
- Unlimited concurrent workers
- Stategraph Cloud or self-hosted, same price
- Priority support, email and Slack
- Onboarding session
- Plans and applies that stay fast at scale
- No global lock
- Queryable record of every change
- Cloud, self-hosted, air-gapped, or BYOC
- Dedicated support and a custom SLA
- Priced separately from Orchestration
Stategraph Cloud vs. HCP Terraform
Both products let you manage Terraform as a SaaS. The difference is what they do with your state.
The key benefits of Stategraph Cloud
Infrastructure provisioning that runs at the pace of your changes, not the pace of your state file.
Your state stays in your database
Every paid tier runs on a single-tenant Postgres instance. Your state is never co-mingled with another customer's infrastructure.
No infrastructure to run
Stategraph Cloud manages the server, upgrades, and availability. You point your Terraform CLI at it and ship.
Start in minutes
Import your existing state with one command. Replace terraform with stategraph. No code changes required.
Plans stay fast as you grow
Stategraph's subgraph execution means plans scale with the size of your change, not the size of your state.
Governance ships with the product
Policy enforcement, audit trails, access control, and drift detection are part of the platform, not add-ons bolted on from the outside.
One system of record
Resources, changes, cost context, compliance evidence, and agent activity all live in the same Postgres model. The scavenger hunt through CI logs ends here.
Ready to get started?
Start on Pro free for 30 days, no credit card, and drop to the Free plan after rather than losing access. Import your existing state with one command, explore the graph, and see how fast plans can be. If you manage Terraform at scale and want faster plans, parallel applies, and cross-state transactions, this is the loop worth trying.
30 days of Pro · No credit card · Free plan after