CockroachDB

A distributed, Postgres-compatible SQL database designed for horizontal scale and multi-region resilience, offered as a managed serverless or dedicated service.

Visit CockroachDB →

Some links may be affiliate links — we may earn a commission if you sign up, at no extra cost to you. It never affects our rankings. Learn more.

CockroachDB is a distributed SQL database that speaks the PostgreSQL wire protocol while running on a shared-nothing cluster of symmetric nodes. Data is split into ranges and replicated (three copies by default) via Raft consensus, providing serializable, strongly consistent ACID transactions that survive node, rack, and datacenter failures. It scales horizontally and can span multiple regions, and is offered both as a managed cloud service and as self-hosted software under a source-available enterprise license.

Key features

PostgreSQL wire-protocol compatibility (SQL, joins, transactions, stored procedures)
Horizontal scale-out for reads and writes by adding symmetric nodes
Automatic range-based replication via Raft consensus (three replicas by default)
Serializable, strongly consistent ACID transactions
Multi-region deployment with data domiciling and zone/region survivability
Distributed vector indexing with pgvector compatibility, plus changefeeds/CDC

Specs

Category	Managed database
Engine	Distributed SQL (Postgres-compatible)
Pricing model	usage-based
Free tier	Yes
Regions	multi-region
Self-hostable	Yes
Open source	No

Last reviewed 2026-07-16 — see our methodology. Specs, pricing and free tiers change often, so confirm current details on the provider before deciding.

Pros & cons

Our editorial assessment — not vendor-stated facts.

Pros

Survives node, rack, and datacenter failures automatically, without manual failover
Scales transactional throughput and storage by adding nodes, unlike single-primary Postgres
Familiar PostgreSQL surface eases migration for many existing applications
Combines strong consistency with geo-distribution, reducing custom sharding logic
Offered as a fully managed cloud service (on AWS, GCP, and Azure) or self-hosted

Cons

Not fully PostgreSQL-compatible: narrower extension ecosystem and some unsupported features and behaviors
Distributed consensus adds write latency, especially for cross-region writes, versus a single-node database
Self-hosted is no longer OSI open source; it uses a source-available enterprise license (free only below a revenue threshold)
Running and paying for a distributed cluster is heavier and often costlier than a single-node database for small workloads

Pricing

Managed CockroachDB Cloud offers a usage-based, serverless Basic tier (billed on Request Units and storage, scales to zero) with a perpetual monthly free allowance, plus provisioned, vCPU-based Standard and Advanced tiers; new sign-ups receive trial credit. Self-hosted software uses a source-available enterprise license that is free for individuals and companies below a revenue threshold and paid above it.

Who should use CockroachDB?

Teams building always-on, transactional applications that need horizontal scalability, high availability, and multi-region or geo-distributed data with strong consistency, while keeping broad PostgreSQL compatibility. Less suited to small, single-region apps where a simpler single-node database is sufficient.

Frequently asked questions

Does CockroachDB have a free tier?: Yes.
How much does CockroachDB cost?: Managed CockroachDB Cloud offers a usage-based, serverless Basic tier (billed on Request Units and storage, scales to zero) with a perpetual monthly free allowance, plus provisioned, vCPU-based Standard and Advanced tiers; new sign-ups receive trial credit. Self-hosted software uses a source-available enterprise license that is free for individuals and companies below a revenue threshold and paid above it.
Who should use CockroachDB?: Teams building always-on, transactional applications that need horizontal scalability, high availability, and multi-region or geo-distributed data with strong consistency, while keeping broad PostgreSQL compatibility. Less suited to small, single-region apps where a simpler single-node database is sufficient.