With the release of dbt Fusion, dbt Labs has introduced a new engine that promises significant performance improvements and advanced capabilities over the familiar dbt Core. Fusion is more than just a technical upgrade: it represents a fundamental shift in how dbt is developed, distributed, and licensed.
In this post, we break down the pros and cons of dbt Fusion from two key perspectives: data teams who rely on dbt for day-to-day analytics, and the open source community that has helped dbt Core thrive.
What is dbt Fusion?
dbt Fusion is a new engine, written in Rust, designed to eventually be a superset of dbt Core. It offers:
- Faster execution than Python-based dbt Core
- Native understanding of SQL across dialects
- Inline SQL linting, CTE previews, and traceability features in the dbt VSCode extension
- Smarter orchestration with column-level lineage and state-aware builds
Fusion is not a simple upgrade to Core. It’s a new piece of software with its own licensing model: a mix of open source, source-available (under the Elastic License v2), and proprietary components.
Data Teams
Pros
- Fusion significantly reduces the time required to run large dbt DAGs, especially helpful for enterprise-scale projects.
- Integrated features in the dbt VSCode extension like inline SQL validation, CTE previews, and model tracing make development more efficient and less error-prone.
- Fusion can rebuild only the models that require updates based on data freshness and column-level lineage, reducing compute costs.
- Most teams using dbt for internal analytics can continue doing so without cost.
Cons
- Some features in Fusion are behind a license key and only available to paying dbt Cloud customers.
- Fusion cannot be used to build managed or hosted dbt offerings without a commercial agreement with dbt Labs.
- Teams may face confusion or inconsistency between dbt Core and Fusion capabilities during the transition.
Open Source Community
Pros
- Much of the Fusion engine remains source-visible, allowing developers to read and learn from the code.
- Fusion allows community contributions after signing a Contributor License Agreement, similar to many large-scale open source projects.
- Fusion is where dbt Labs is investing in new features and performance improvements.
Cons
- ELv2 limits how the software can be used, especially in commercial/hosted offerings.
- Unlike Apache-2-licensed dbt Core, Fusion includes proprietary parts, which hinders complete transparency and free redistribution.
- Third-party projects, especially SaaS tools and adapters, will have more limited integration opportunities.
- The presence of proprietary components muddies the waters on how much influence the community has over Fusion’s direction.
A Summary of Key Licensing Differences
| Feature | dbt Core | dbt Fusion |
|---|---|---|
| License | Apache 2 | Elastic License v2 + proprietarys |
| Source Code | Fully open | Mostly visible, partly proprietary |
| Use in Business | Fully open | Free for internal use only |
| Contributions | Fully open | Allowed with agreement |
| Hosted Services | Permitted | Not allowed without a license |
| Ecosystem Impact | Encourages plugins & adapters | Some limits for third-party tools |
Final Thoughts
dbt Fusion is the future of dbt’s engine, but it’s not the same future that open source advocates may have hoped for. For data teams, it brings faster runtimes, better tooling, and smarter orchestration. For the open source community, it’s a more mixed picture: Fusion is visible and extensible, but not fully open.
If you’re a business just using dbt internally, Fusion is a welcome upgrade. But if you’re building products or services on top of dbt, the new licensing model demands careful consideration.
As Fusion develops, it will be important to watch how dbt Labs balances commercial strategy with community openness. For now, dbt Core remains a powerful, open, and unrestricted tool for analytics engineering, but Fusion is where the next wave of features will land.