Transformations Pricing FAQ

Read answers to frequently asked questions about Fivetran Transformations pricing updates on January 1, 2025.

When will Fivetran start billing for transformations?

Fivetran has started billing for transformations as of January 1, 2025. This includes usage pertaining to Quickstart data models and Fivetran-hosted Transformations for dbt Core.

NOTE: We are not charging for usage related to the Fivetran Platform Quickstart data model.

Do I need to pay for third-party orchestrations like dbt Cloud™ Integration and Coalesce?

Third-party orchestrations are free on all pricing plans.

How does transformations pricing work?

We calculate transformations usage based on the total number of successful monthly model runs. Failed runs will not count towards the monthly total. Pricing is usage-based, and the higher the usage, the lower the cost per successful model run.

See the transformations pricing tiers and example usage below:

NOTE: We’re offering up to 5,000 model runs per month for free to help customers manage costs. This applies to all Fivetran pricing plans.

What are monthly model runs?

Monthly model runs are the number of models successfully executed in a given month. Fivetran uses it as a metric to calculate transformations activity and usage levels.

What is a model run?

A model run is an SQL statement that produces a materialized table or data view. Transformations often consist of multiple interdependent models that work together to produce the final dataset used for business intelligence (BI) or operational purposes.

Are both successful and failed model runs counted as usage?

We count only successful model runs as usage. Model runs that fail do not count toward your usage.

If a job is partially completed, do the successful model runs still count toward paid usage?

If you have a transformations job that has only been partially completed, any successful model runs that are part of that job count towards your total usage.

Does re-running the same model within a month count multiple times toward my model run total?

Yes, each successful model run counts towards your usage. For example, if you have a job that runs five times a month, for each of those runs, all successful model runs count towards your total usage.

If a job has only been partially completed and I need to restart it, will the previously successful model runs be counted again?

Yes, we count every successful model run for each job separately. Even if a job is only partially completed, all successful model runs that are a part of that job still count toward your total usage. When you restart the job, we count all successful model runs separately.

Do I need to do anything before January 1, 2025?

We encourage you to review your existing jobs. You may have jobs that are outdated, unnecessary, or failing consistently. Removing jobs you don’t need optimizes your usage, so you don’t pay for model runs you don’t need. See our usage optimization video guide for more details.

We're also updating our Service Consumption Table to add Transformations and to simplify some of the existing Service Consumption Table language. Using Transformations on or after January 1, 2025 will constitute acceptance of these updated terms and associated pricing.

I’m on the Free plan. Will I be able to use Fivetran Transformations?

The Free plan still supports Quickstart data models and dbt Core Transformations, but usage is limited to 5,000 monthly model runs.

What happens if I exceed the 5,000 monthly model runs limit on the Free plan?

If you exceed the 5,000 monthly model runs limit on the Free plan, you can either opt to optimize your usage or upgrade to a paid plan. The first time you exceed the limit, your transformations will continue to function during the next billing cycle.

If your transformations usage exceeds 5,000 monthly model runs for the second time:

• Fivetran will pause your connectors and transformations jobs and freeze your account.
• To continue using Fivetran, you would need to unfreeze your account for 24-hours and upgrade to a paid plan in your dashboard.

How can I track my monthly model runs?

The Billing & Usage page in your dashboard provides transformations usage estimates based on historical usage data in your account. There, you can see the usage breakdown by day, the jobs run, and the associated usage for each job.

How can I optimize my total monthly model runs to minimize costs?

You can control costs by reducing the number of monthly model runs. See our Transformations pricing best practices video below for practical advice.

You can optimize your costs by adjusting your scheduling to change the run frequency or by only selecting Quickstart output models that matter the most.

Custom scheduling

Custom scheduling allows you to select particular days of the week to run your transformation, with a run frequency within a day from 1 minute to 24 hours. If you choose 24 hours, you can set a particular time of day in 30-minute increments. Reducing your run frequency reduces your monthly model runs and costs. Consider how up-to-date your transformations need to be at all times.

Integrated scheduling and connector sync frequency

If you are using integrated scheduling, adjust it to optimize the number of monthly model runs. Integrated scheduling automatically triggers model runs following the completion of Fivetran connector syncs, so transformations run only when there’s new data. In this case, the usage depends on how often your connectors sync. To optimize your total monthly model runs, adjust the sync frequency of your connectors.

Cron jobs

Cron jobs are tasks that run at specific times or intervals. For example, you can schedule your dbt models to run hourly, daily, or weekly based on your reporting or analytics needs. We use them in our deployment.yml-based scheduling for all our transformations. Cron jobs will also be available in the dashboard on January 1, 2025.

Quickstart data models

You can also control costs by selecting the Quickstart output models that matter most to you. Each of our Quickstart data models has multiple output models associated with it. Output models are the final model in a transformation workflow. Similar to selecting tables within a schema for connectors, you can choose which output models you need with Quickstart. If certain output models are not valuable, you can turn them off.

How do I know how many individual models are a part of a Quickstart?

The number of models associated with each data model is available in Fivetran’s dbt packages documentation. See our Available Data Models documentation and go to the relevant data model’s dbt package documentation.

The Quickstart data model I use produces five output models, but I am being charged for more. Why is that?

While output models are ready for use, a transformation often requires more models in intermediate transformations to produce the final output models. These are known as intermediate models. As a result, the total number of models includes many intermediate models, which you do not see but are essential for producing the final output models.

Can I run Fivetran’s data models outside of Fivetran? Will I be charged for that?

All Fivetran data models are open-source, and customers can download them from the dbt Hub and orchestrate them outside of Fivetran. We only charge for the simple, no-code setup and maintenance that Fivetran takes care of for customers. If you prefer, you can run the models outside of Fivetran and not use Quickstart.

How can I turn off my transformations?

You can remove your transformations in your Fivetran dashboard. To do so:

1. Log in to your Fivetran dashboard.
2. Click Transformations and select the destination for which you want to turn off transformations jobs.
3. Select the transformations you want to remove by selecting the checkboxes beside their names.
4. Click Remove.
5. In the confirmation pop-up window, click Remove.

* dbt Core is a trademark of dbt Labs, Inc. All rights therein are reserved to dbt Labs, Inc. Fivetran Transformations is not a product or service of or endorsed by dbt Labs, Inc.