Starburst Galaxy

Overview

Starburst Galaxy is a managed cloud analytics service built on Trino for querying data across systems with SQL. It suits organizations that want to analyze distributed data without taking on complex cluster management. Its combination of federation and Iceberg upkeep is useful, but usage-based charges make compute and ingestion costs important to weigh.

Key features

Galaxy can query data lakes and warehouses together in one SQL query without copying data. It connects to more than 50 sources, including Amazon S3, Snowflake, Redshift, BigQuery, and Azure Blob, and is available on AWS, Google Cloud, and Microsoft Azure. That breadth fits teams with data spread across these systems; a team working from a single source may not need a federation layer.

Galaxy can stream data from Kafka or batch ingest CSV, JSON, and AVRO into Iceberg. LakeOps automates Iceberg compaction, snapshot expiry, and orphan-file removal on a dedicated serverless engine, reducing routine table maintenance. Ingestion compute is billable, however, so this operational convenience has a usage cost.

AIDA turns natural-language questions into governed insights through a customizable interface or MCP. Its token usage is billed separately on Enterprise and Mission-Critical, a cost to factor in for teams planning to use it regularly. Security features include fine-grained RBAC and ABAC, end-to-end encryption, AI-powered data classification, and private connectivity. Starburst describes Galaxy as SOC 2 Type II certified and HIPAA-ready.

Galaxy supports SQL analytics, ACID transactions, open table formats, streaming ingestion, and a governance catalog. Users can contact Starburst support professionals from within the product. Its REST API runs over TLS/HTTPS, with authentication tokens that expire after 10 minutes.

Pricing

Galaxy has a free plan and usage-based paid plans. The headline rates are credit rates rather than a complete cost estimate: documented billable activity includes cluster compute hours, data ingestion compute, and private-link data transfer, with rates determined by plan.

• Free: free, billed Free forever. It allows up to three clusters and standard execution mode for ad hoc queries. This is a useful way to begin with limited workloads, but the cluster cap and execution mode constrain its role for broader production needs.
• Pro: price on request, billed Starting at $0.50/credit. It adds flexible cluster execution modes, streaming ingestion, and advanced cluster management, making it the step up for teams needing those capabilities.
• Enterprise: price on request, billed Starting at $0.75/credit. It adds advanced autoscaling, ABAC and SCIM, and AWS PrivateLink for data sources. AIDA token usage is billed separately.
• Mission-Critical: price on request, billed Starting at $1.00/credit. It adds elite support, advanced governance integrations, and the highest uptime guarantees. AIDA token usage is billed separately.

A 30-day free trial includes Enterprise features and $500 in Starburst Galaxy compute resources. That gives teams a defined allowance to assess higher-tier capabilities; it is distinct from the Free plan, whose three-cluster cap applies indefinitely.

Platforms

Galaxy is available through web and API, with Linux also listed as a platform. Its cloud availability spans AWS, Google Cloud, and Microsoft Azure.

Who it's for

Galaxy is a strong fit for organizations that want federated SQL across cloud data sources and managed Iceberg ingestion and maintenance without building the expertise to run complex data clusters. It is less suitable for teams that need predictable costs without monitoring usage, require self-hosted deployment, or only query a narrow set of data.

Pros and cons

• Pro: Federated SQL can query lakes and warehouses together without copying data, useful when information is distributed across many systems.
• Pro: Ingestion and automated Iceberg maintenance bring data landing and table upkeep into the same managed service.
• Pro: The free tier and a trial with Enterprise features provide two ways to evaluate Galaxy, including a trial compute allowance.
• Con: Usage charges span compute, ingestion, and private-link transfer, so the credit rate alone does not determine a workload's total cost.
• Con: The Free plan is limited to three clusters and ad hoc queries in standard execution mode.
• Con: AIDA adds separately billed token usage on Enterprise and Mission-Critical.

Alternatives

Oracle Autonomous AI Lakehouse is another freemium option with an unlimited-time free usage plan subject to capacity limits; consider it if that free-plan structure better fits your evaluation.

Starburst Data Platform also offers a free plan capped at three clusters and a Pro plan starting at $0.50/credit; compare it if you are weighing another Starburst offering with those tiers.

Amazon SageMaker Autopilot is a paid service with a free trial and pay-as-you-go pricing; consider it if that pricing model suits your needs.

Apache Hudi is a free, self-hosted option for teams seeking an open-source lakehouse platform rather than a managed cloud service.

Bauplan offers a freemium sandbox with public datasets or uploads, a CLI, SDK, API, and community support; consider it if a shared sandbox suits your work.

Databricks Notebooks offers a free edition with one serverless workspace and limited compute size and usage; consider it if notebook-based work within those limits is a better fit.

Google Cloud Lakehouse for Apache Iceberg is a paid option with a metadata-storage free tier of 1 GiB per month; consider it if that scope matches your requirement.

IOMETE offers a free self-hosted, on-premises plan capped at 100 vCPUs; consider it if deployment control is more important than managed cloud operation.

Galaxy also sits in Data Lakehouse Platforms, Query Engine Software, Data Federation Software, and Data Virtualization Software.

Verdict

Choose Starburst Galaxy if your organization needs SQL across distributed cloud data and wants ingestion and Iceberg maintenance managed alongside it. The free tier and Enterprise-feature trial lower the barrier to evaluation, while its broad connectors address a real multi-source need. Look elsewhere if you need tightly predictable costs, self-hosting, or a simpler setup than a multi-source analytics platform.