Announcing dbt Core v2.0 and Fusion: two distributions powered by a single engine

On May 28, 2025, dbt Labs announced the next-generation dbt Fusion engine, which is faster, more capable, and more efficient than the Python engine that powers dbt Core v1.x.

On June 1, 2026, dbt Labs announced that the dbt Fusion engine will power both the dbt Core (OSS) and Fusion (proprietary) distributions of the dbt v2.0 framework, and that the corresponding code has been published to the dbt-core repository under the Apache 2.0 license.

Is the license for dbt Core v1.x changing?

No, it is still Apache 2.0.

Is the license for dbt Core v2.x changing compared to v1.x?

No, it is Apache 2.0 as well.

Will dbt Labs continue to support and maintain dbt Core?

Yes.

What is the difference between the dbt Core and Fusion distributions?

Version 2 of the dbt framework has two distributions which can both be installed locally for free, powered by a single engine:

• dbt Core is completely open-source. Its code and binary are subject to the Apache 2.0 license.
• dbt Fusion extends dbt Core with additional proprietary code to create an enhanced binary with additional features.

Teams (and even other competitive vendors) can use the Fusion distribution for free, subject to a handful of basic provisions described below. Some additional features of the dbt Fusion distribution require a dbt platform account or paid subscription.

We recommend that most teams install the dbt Fusion distribution, which has more capabilities available than its Apache counterpart even if you never engage with dbt Labs.

What happened to the ELv2-licensed code in the dbt-fusion repository?

It is now in the dbt-core repository under the more permissive Apache 2.0 license.

The additional code necessary to build a Rust implementation of dbt Core – which we previously committed to releasing under the Elastic License (ELv2) by the time the dbt Fusion engine reached General Availability – has also been moved into the dbt-core repository as Apache 2.0 code.

Do I need a dbt platform account, or to be a paying dbt platform customer in order to use dbt Fusion?

No.

Do I need a dbt platform account, or to be a paying dbt platform customer in order to access some premium features of dbt Fusion?

Yes.

Is there any limit to the number of users at a company who can use the dbt Fusion CLI for free?

No.

What do I, a dbt Core user, need to know about dbt Fusion?

As long as you aren't offering dbt Fusion as a hosted or managed service to a third party, you should think about using dbt Fusion in almost exactly the same way you think about using dbt Core today. It’s free, you can contribute to its development, and you can adopt it across your organization without asking permission or changing your existing deployment method.

You can use dbt Fusion internally in your own business, for free and without restriction. This includes using dbt Fusion to provide transformed data to other customers. However, if you are providing your customers a product with dbt Fusion embedded, you must permit your customers to use the full features of dbt Fusion, including login-gated features if they choose.

Licenses

What licenses does dbt Labs use for dbt products?

dbt Labs uses different license types for different software offerings:

• dbt Core is released under the Apache 2.0 license, which is an Open Source software license that permits anyone to use, copy, redistribute, or modify the licensed software without warranty, conditions, or limitations, aside from the requirement to preserve license notices, attribution, and trademarks.
• dbt Fusion is proprietary to dbt Labs and extends the capabilities of dbt Core. dbt Fusion is made available to customers and partners under the dbt Product Licensing Agreement, which allows broad adoption subject to terms designed to support the sustainable business of its developer and maintainer (dbt Labs).
• Other software is proprietary to dbt Labs and made available to customers and partners under commercial terms of service. This includes software that is available for free, within usage limits.

Understanding dbt Core and dbt Fusion

Has the license for dbt Core changed?

No. There is no change in license for existing dbt Core software, or future releases of that software.

dbt Core (which includes the dbt-core, dbt-common, and dbt-adapters repositories maintained by dbt Labs) is licensed under Apache 2, as it has been since its first release in 2016.

As Fivetran + dbt Labs President Tristan Handy wrote in a blog post introducing the dbt Fusion engine, and again when announcing Fivetran and dbt Labs’ merger, the combined company will continue maintaining dbt Core under the Apache 2 license indefinitely.

dbt Labs is committed to the ecosystem around dbt Core, and the thousands of users and customers who rely on it for their data work every day.

Does a single shared engine mean dbt Core is being replaced?

No.

dbt Core is not:

• a specific set of Python files.
• the version of dbt you can run on a laptop.
• the version of dbt you can run without paying.

dbt Core is the completely open source implementation of the dbt framework.

The dbt framework includes a language and an engine.

The dbt language, the code you can write in your dbt project, has become a standard for the industry. dbt Labs has developed this language, with contributions and input from the community, in dbt Core over the past decade.

The dbt engine is the foundational technology for compiling dbt projects, executing transformation graphs, and producing metadata. There is an entirely new next-generation engine for v2 of the dbt framework, written in a different programming language (Rust) than dbt Core v1.x (Python).

Instead of maintaining two engines in two languages, dbt Labs has created a single engine which powers two distributions by open sourcing the code required to produce a Rust distribution of dbt Core v2. We are committed to maintaining and continuing to expand the dbt language across both the Core and Fusion distributions, exactly as we have done to date. You should expect to see new dbt Core versions (v2.1, v2.2, etc) that include support for new language features and fixes. dbt Labs will also continue to evaluate and integrate external contributions from the community.

dbt Labs’ goal is to enable as many users as possible to upgrade to Fusion when they adopt version 2 of the dbt framework. We believe that Fusion will provide a superior development and deployment experience for everyone.

You do not need to be a paying dbt customer to use Fusion.

How is dbt Fusion licensed?

The dbt Product Licensing Agreement terms only apply to the dbt Fusion binary and strike the balance of providing openness and flexibility for users, while also providing protections for dbt Labs to build a sustainable business as the software’s primary developer and maintainer.

As a user of dbt Fusion, you can do almost anything with the software, except:

• If you redistribute dbt Fusion to others as a managed service, you must permit your end users to utilize the full capabilities of dbt Fusion, including features gated behind login, if they choose to do so. User login must be accessible through standard documented flows.
• You cannot circumvent dbt Fusion’s “license key” functionality or attempt to derive Fusion’s source code.
• You cannot remove or obscure any notices about dbt Fusion’s license or dbt’s trademarks/copyrights.

Why does dbt Fusion not use ELv2 any more?

The primary purpose of the Elastic License (ELv2) is to prevent a tool’s use in competing products which host that tool as a service. Now that dbt Core v2.0 is built on the same engine – using code which was relicensed from ELv2 to Apache 2.0 – preventing competitive deployments of dbt Fusion is not a priority.

Third party vendors can offer dbt Fusion as a managed service – including the advanced features which are available without login – provided they do not prevent users from logging in and enabling other premium features. This also aligns with the goals of Open Data Infrastructure, which at its heart prioritizes flexible tool choice, low switching costs, and interoperability with established standards.

We would prefer users building on top of third-party tools (e.g. a different orchestrator) to adopt the full dbt Fusion binary, so they can seamlessly opt into additional services from dbt Labs in the future if they choose to.

What’s the difference between dbt’s public source code and binary distributions?

The Core and Fusion distributions of dbt v2 are both derived from a single engine which is written in Rust, a compiled language. dbt Core’s source code is visible in the dbt-core repository, and also distributed as a compiled binary.

You are welcome to read, compile, modify, and redistribute the visible source code and the accompanying dbt Core binary. Both are licensed under the Apache 2.0 license.

dbt Fusion is based on the same Apache-licensed source code in the dbt-core repository, and also includes proprietary components for which the source code is not publicly viewable. These components add functionality, some of which is available to everyone, and some which is available exclusively to logged-in or paying customers (via license key). Our documentation site will make it clear when functionality requires the proprietary Fusion distribution.

The dbt Fusion distribution is subject to the dbt Product Licensing Agreement as described above.

The dbt language spec and adapter functionality are freely available in the dbt-core repository and do not require a license key.

What is the “license key” within the context of dbt Fusion?

Some of dbt Fusion’s premium features are proprietary functionalities that are available only to registered or paying customers, and are unlocked via a license key. dbt Fusion’s license key is implemented via authentication with dbt Labs’ commercial software. By authenticating, users unlock these additional features.

The dbt Product Licensing Agreement prohibits any attempt by any user to circumvent license key functionality.

Does dbt Labs welcome contributions to dbt Core and dbt Fusion?

Yes! dbt Labs maintains many open-source software projects in public repositories on GitHub (https://github.com/dbt-labs). Anyone can contribute code to any of these projects after signing our Contributor License Agreement. If you are a user of dbt Fusion, contributions you make to the dbt-core repository will apply to both distributions since they share the same underlying code.

As always, we value the participation of the community, and will carefully review and evaluate external contributions. We also recognize that valuable contributions come in many forms: opening issues, commenting in discussions, answering questions from other community members, and contributing code to fix bugs or extend functionality.

Can I use dbt Fusion?

Examples by use case

My team is using dbt to develop, transform, and test data assets in my company's cloud data warehouse. Those data assets are powering real-time dashboards for our colleagues and embedded product experiences for our customers. Is this a permitted use of dbt Fusion?

Yes, this is permitted. We believe strongly in the “dbt way” of doing analytical work (which we’ve called the ADLC), and we believe that upgrading to dbt Fusion will enable your team to do this work even more effectively.

I'm a contractor or consultant setting up dbt Fusion for my clients to use internally. Am I permitted to continue doing this? Am I subject to the limitations on redistribution?

Yes, this is permitted and is not considered redistribution of dbt Fusion, assuming each client’s deployment is standalone and they could control it directly if they wanted to.

I want to use dbt Fusion to transform my customers’ data into assets that power my application. The app shows Fusion data test failures directly to customers, as a way of alerting them to data quality issues in their source data systems. Am I permitted to continue doing this?

Yes, this is permitted, because your customers are not using a tool that embeds and redistributes dbt Fusion directly. Rather, your customers are accessing the data itself, which is transformed using dbt Fusion behind the scenes.

I’m building a general-purpose data preparation and analytics product. I provide my customers with data transformation workflows that are leveraging dbt Fusion without exposing the dbt Fusion login mechanism to my customers. Is this a permitted use?

No.

If your customers are submitting code, configuration, or other inputs to data transformation workflows embedded in your product, and your product is powered by dbt Fusion, users must be able to create an account and login to enable dbt Fusion’s premium features directly. It does not matter whether your customers’ inputs happen via UI, API, prompts, or otherwise.

However, you can use dbt Core v2.x in any way you choose, or modify the code in the dbt-core repository to meet your needs. If you take this approach, you must still respect dbt Labs’ trademark policy and brand guidelines.

If you would like to discuss your specific case, and the possibility of a negotiated commercial agreement for dbt Fusion, please reach out to us at license@dbtlabs.com.

I’m building a data product that provides functionality complementary to dbt. I provide my customers with a dbt Fusion integration as part of my hosted/managed service. Is this a permitted use?

Yes, provided you comply with the dbt Product Licensing Agreement.

If your product integrates with documented APIs of dbt — such as by generating dbt code to write back to your customers’ projects, or by parsing metadata artifacts that your customer provides from their own self-hosted dbt Fusion deployment — this is permitted without reservation (and encouraged!). dbt Labs publishes validation schemas and artifact specifications with Apache 2 licenses, which you may freely use as part of building your integration.

If your product integration invokes dbt Fusion against customer projects as part of your hosted or managed service, this use is permitted as long as you comply with the Redistribution requirements in the dbt Product Licensing Agreement. For example, you cannot block a customer’s network access to dbt servers, because they must be able to enable premium functionality if they choose to do so. Please reach out to us at license@dbtlabs.com with any questions.

Examples by deployment pattern

If I am self-hosting Airflow to run dbt Core for my own business today, can I start running dbt Fusion instead?

Yes.

If I am an end user of a hosted/managed service that provides dbt Core today, can they offer dbt Fusion to me directly as a feature of their service?

Yes. The hosted/managed service provider must not prevent you from enabling Fusion’s premium features if you choose to do so, but they may provide you with access to dbt Fusion (which has additional capabilities beyond dbt Core even if you do not log in).

If I am an end user of a hosted/managed service that provides dbt Core today and has not implemented native dbt Fusion support, am I in violation of the license if I install dbt Fusion directly?

No, this is not a violation of the license. As an end user of dbt Fusion, you are not subject to the limitations applied to redistributors of Fusion.

If you are the provider of a hosted/managed service that might run dbt Fusion, or if you have a different question not answered here, we encourage you to contact us at license@dbtlabs.com.

If I currently deploy dbt Core in an airgapped environment with no external network access, am I allowed to replace it with dbt Fusion?

Yes. You are allowed to deploy dbt Fusion into any environment you choose, with any networking restrictions required by your business. The dbt Product Licensing Agreement only prevents a third party from blocking your ability to enable Fusion’s premium features if you choose to do so.

If I currently deploy dbt Core to developers across my company with preconfigured defaults which disable anonymous usage stats, am I allowed to replace it with dbt Fusion?

Yes. The Redistribution restrictions in the dbt Product Licensing Agreement only apply to those providing dbt Fusion to third parties.