What is CloudQuery?

CloudQuery is a cloud asset inventory designed for platform teams. It pulls configuration and security data from AWS, Azure, GCP, and over 70 cloud and SaaS providers, normalizing everything into a single schema. The synced data lands in PostgreSQL, BigQuery, Snowflake, S3, or other destinations you run. You declare sources and destinations in version-controlled YAML and execute syncs on your own infrastructure — as a standalone binary, through Docker, or via Homebrew. The tool uses Apache Arrow for fast incremental syncs across hundreds of accounts. An MCP server lets AI agents query your cloud estate directly with SQL. The CLI and framework are open source under MPL-2.0; some plugins have moved to closed source but remain available under MPL-2.0 in the git history. Pricing is based on rows synced per month, with a free tier up to 1 million rows.

Should you self-host CloudQuery?

CloudQuery scores 65/100 on selfhost.directory (how scoring works) and ranks #3 of 7 in Data Pipelines. It has 6 545 GitHub stars.

It shipped 8 stable releases in the past year. The latest, plugins-destination-kafka-v5.8.3, came out on 22 October 2026.

Setup is moderate: expect some configuration beyond a single command. An official Docker image is available. It is MPL-2.0-licensed and written mostly in Go.

Its closest alternative on selfhost.directory is vector (79/100, 22 777 stars). It scores higher overall, so compare the two before you pick. See CloudQuery vs vector side by side.

Features

Apache Arrow performance
Cloud asset inventory
Cloud FinOps
Cloud Security Posture Management (CSPM)
Extensible plugin system
Historical snapshots
Incremental sync
MCP server for AI agents
Multiple destination support
Policies
Reports and dashboards
Self-hosted data privacy
SQL querying
Version-controlled YAML configuration

FAQ

What if CloudQuery doesn't support a source I need?

Build it yourself with the Go, Python, JS, or Java SDK, or check the 80+ existing connectors first.

Does CloudQuery support AI agents or MCP?

Yes, connect an MCP server to your synced data so AI agents can query it directly. The MCP server runs in PostgreSQL or Snowflake mode without a Platform subscription.

Which databases can I sync data into?

PostgreSQL, BigQuery, Snowflake, S3, and other supported destinations configured in YAML.

Can I self-host CloudQuery entirely on my own infrastructure?

Yes, CloudQuery is self-hosted by design and runs on infrastructure you control, syncing to a destination database you choose.

Is CloudQuery free to use?

Yes, CloudQuery is open source and free to run. You only pay for your own infrastructure plus usage-based billing for data synced if using a paid destination.