Mage vs. Fivetran

First published on November 17, 2022

Last updated at May 2, 2023

 

8 minute read

Tommy Dang

TLDR

Mage is an open-source data pipeline tool for transforming and integrating data. Find out how Mage compares to ELT alternatives like Fivetran. We’ll break down the features, pricing, pros, cons, and more.

Outline

  • About the products

  • Feature comparison

  • Next steps

⁠About the products

Mage

is an open-source data pipeline tool for integrating (EL) and transforming (T) data. Mage was started in 2020 by engineers who worked on data and dev tools (e.g.

) at Airbnb for 5+ years.

Fivetran

is a SaaS (software-as-a-service) company providing a managed ELT service. Their enterprise software is closed-source. Fivetran was founded in 2013.

Feature comparison

Mage

Fivetran

Specialty

Data pipelines

Data ingestion

Number of sources

307 (from Singer taps and targets)

160

Number of destinations

20 - all the main ones and more (from Singer taps and targets)

All the main ones

Customizations

Mage uses the data engineering community standard for data integrations called the Singer Specification. All connectors are written in Python.

⁠Mage provides tutorials, guides, examples, and training to create custom connectors.

Write code in Fivetran's Cloud Functions.

Database replication

Full table and incremental via CDC (change data capture)

Full table and incremental via CDC (change data capture)

Integration with modern data stack

Integrate with any Python library.

⁠Mage has a native integration with DBT: preview DBT results, orchestrate DBT model runs, schedule DBT models to depend on non-DBT tasks (e.g. ETL/ELT pipelines).

DBT

Support and developer documentation

Support is instantly provided through Slack (

) and via email (

)

1-on-1 tech support is provided over Zoom and can easily be schedule through Slack and email

Documention (

) has been praised by data engineering community

Support is provided through a form and someone follows up; no instant support through chat

Documentation is really good, but no community via Slack or Discord

Fivetran offers no training services

Support service-level agreements (SLA)

Mage is self-hosted

Available

Security

SOC2

HIPAA, GDPR, SOC2

Vendor lock-in

Mage is open-source. The code and connectors you use and write are always yours and available to you if you switch tools.

Mage's technical design makes your code and connectors modular and interoperable.

Pay month-to-month or lock into annual contracts.

Closed-source.

Pricing

Free (self-hosted on AWS, GCP, or Azure)

Usage based pricing. Costs are based on monthly active rows (e.g. rows inserted, updated or deleted by Fivetran connectors).

Specialty

Mage focuses on data pipelines while Fivetran provides ELT services for data ingestion.

Number of sources

Mage currently offers 307 data source connectors while Fivetran provides around 160.

Number of destinations

Both Mage and Fivetran can add all the main destinations, while Mage also provides destinations from Singer taps and targets.

Customizations

Mage uses the data engineering community standard for data integrations called the Singer Specification. All connectors are written in Python. Mage also provides tutorials, guides, examples, and training to create custom connectors.

With Fivetran, you can write code in Fivetran's Cloud Functions.

Database replication

You can replicate full table and incremental via CDC (change data capture) for both Mage and Fivetran.

Integration with modern data stack

With Mage, you can integrate with any Python library. Mage has a native integration with DBT:

  • Preview DBT results

  • Orchestrate DBT model runs

  • Schedule DBT models to depend on non-DBT tasks (e.g. ETL/ELT pipelines)

Fivetran integrates with DBT.

Support and developer documentation

Mage provides instant, real-time support via Slack (

) and via email (

). Mage also provides 1-on-1 tech support or group training services over zoom which can be easily scheduled through Slack or email. Last but not least, Mage's documentation (

) has received some recognition in the data engineering community.

Fivetran offers support through a form in which you can submit and impressive documentation. Unfortunately, there is no instant support, community, or training services available.

Support service-level agreements (SLA)

Both have SLAs available, while Mage is self-hosted.

Security

Both are SOC 2 compliant, while Fivetran is also HIPAA and GDPR compliant.

Vendor lock-in

Mage is open-source. The code and connectors you use and write are always yours and available to you if you switch tools. Mage's techincal design makes your code connectors modular and interoperable.

With Fivetran, you would pay month-to-month or be locked into an annual contract. They are also closed-source.

Pricing

Mage is free as long as you are self-hosted (AWS, GCP, or Azure).

Fivetran offers usage based pricing. Costs are based on monthly active rows (e.g. rows inserted, updated or deleted by Fivetran connectors).

Next steps

Want to learn more? Check out the

.

Want a demo from a real human?

call with our team.

Want to chat? Join our

.