The best tools to replicate MySQL to BigQuery, compared

Every serious option for keeping BigQuery in sync with MySQL reads the MySQL binary log (change data capture), so they all handle inserts, updates and deletes. They differ in how volume is priced, how quickly changes become queryable in BigQuery, how many other sources they cover, and what you have to run yourself.

Quick answer

• Only MySQL to BigQuery, priced per row, with change history: Quayen.
• Everything inside your Google Cloud project, several database engines: Google Cloud Datastream.
• Many SaaS and database sources in one platform: Fivetran, Hevo Data or Airbyte Cloud.
• Open source you run yourself: Airbyte (self-hosted) or Debezium.
• Real-time movement to several destinations, with in-flight transformations: Estuary.

The tools

Google Cloud Datastream. Google's serverless CDC service for MySQL, PostgreSQL, Oracle, SQL Server and more, writing to BigQuery or Cloud Storage. Priced per GiB processed (CDC from $2.00/GiB for the first 2,500 GiB a month). A natural choice if you want everything in Google Cloud billing. Quayen vs Datastream.

Fivetran. A broad integration platform with hundreds of connectors, priced by monthly active rows with a free tier of 500,000 MAR. Strong if you need many sources and a large vendor. Quayen vs Fivetran.

Airbyte. Open-source with a large connector catalog, self-hosted or as Airbyte Cloud (database sources use credits per GB synced). MySQL CDC changes are read on each scheduled sync. Quayen vs Airbyte.

Hevo Data. No-code pipelines for databases and SaaS apps. The Free plan covers 1M events a month; Starter starts at $299/month with hourly syncs, and streaming pipelines need Professional or above. Quayen vs Hevo.

Estuary. Real-time data movement with 200+ connectors, priced at $0.50/GB plus $100/month per connector instance, free up to 10 GB a month. Its BigQuery connector loads on a sync schedule (30 minutes by default). Quayen vs Estuary.

Debezium. The leading open-source CDC engine. It publishes change events, usually to Kafka; turning them into up-to-date BigQuery tables is a sink and merge logic you build and operate. Quayen vs Debezium.

Quayen. A managed service for one path: MySQL (RDS, Aurora, Cloud SQL, Azure, self-hosted) to BigQuery. Changes reach a change-history table within seconds and up-to-date tables refresh every 1 to 60 minutes, with guided setup per host and BigQuery cost controls built in. Priced by monthly active rows and free during early access. Pricing.

What to check before choosing

• Deletes: are rows removed from the BigQuery copy, or only flagged? See syncing deletes.
• Latency to a queryable table: “real time” capture can still land in BigQuery on an hourly or 30-minute schedule.
• How volume is measured: rows changed, events, or GB processed give very different bills for the same database.
• BigQuery costs: a MERGE with updates or deletes is billed for the whole target table it scans; ask how often the tool merges. See keeping BigQuery costs low.
• Binary log retention: any CDC tool needs the binlog kept long enough to recover from an outage. See binlog retention.

Competitor details are from each vendor's public documentation as of November 2026; the comparison pages link the sources.