dbt Documentation (Developer Hub)

What it is

The dbt Developer Hub holds the product documentation, command and configuration reference, best-practice guides, the developer blog and links to the community. It covers the open source distribution you run yourself and dbt's hosted platform.

The pages worth reading first

• Quickstarts: set up a project against DuckDB, BigQuery, Snowflake, Databricks, Redshift or another warehouse.
• Build your DAG: models, sources, seeds and snapshots, and how ref and source connect them.
• Materialisations: view, table, incremental and ephemeral, and when to use each. The incremental models page is essential once your tables grow.
• Data tests: the built-in tests (unique, not_null, accepted_values, relationships), custom tests and unit tests.
• Best practices, especially How we structure our dbt projects, which explains the staging, intermediate and marts layers most teams use.
• Snapshots, for tracking slowly changing dimensions.

Versions in 2026

dbt v2 reached general availability in September 2026. It is built on the engine previously called Fusion; the engine is now called dbt, and the Apache 2.0 open source distribution is dbt OSS. The docs have a version selector: check that the page you read matches the version you installed. Most concepts are the same across versions, but some commands, flags and configuration options changed.

How to use it well

Do not read it cover to cover. Use it the way working engineers do: build something, and when you hit a question (how do I make this model incremental, why is this test failing, how do I document a column) search the docs and read that page fully, including the examples. Then read the best-practice guide on project structure once your project has more than a handful of models.

Prerequisites

Good SQL and a warehouse or local database to run against. DuckDB is a free, easy choice for learning.

Is it worth your time?

Yes. The docs are clear, well maintained and full of runnable examples. They are the most reliable source when tutorials disagree, which happens often with a tool that changes as fast as dbt.

Common questions

Is dbt free? The open source distribution is free under Apache 2.0. The hosted platform has its own plans and prices; check the dbt pricing page for current details.

Is there a community? Yes. The dbt Community Slack and forum are large and active, and linked from the docs.