Databricks Genie is the name Databricks gives to its AI experiences for people who work with data. In the words of the Databricks documentation, it gives every user one place to ask data questions in natural language, explore dashboards and run apps, with every answer grounded in the organization's own data and governed through Unity Catalog. Genie is a family of products, not one feature: Genie One for business users, Genie Agents for data teams, and Genie Code for developers.

Key facts

  • What it does: you ask a question in plain language, Genie writes and runs the SQL, and shows the answer with its sources.
  • Who it is for: business users (Genie One), data teams who set up trusted data (Genie Agents), and developers (Genie Code).
  • Cost: Genie One and Genie Agents are free for human users through January 31, 2027. Genie Code has a free monthly allowance per user and is billed beyond it.
  • Governance: queries run as you, so your Unity Catalog permissions, row filters and column masks apply.

Everything here comes from the Databricks documentation as it read on November 21, 2026. Each product has a longer article on this site, linked where it comes up.

The Genie family at a glance

ProductWho uses itWhat it does
Genie OneBusiness usersA simple chat and home screen for asking data questions and opening dashboards, agents and apps shared with you
Genie AgentsData analysts and data teamsA chat space for one business area, set up with chosen tables, example queries and instructions, so answers are trusted
Genie CodeDevelopers and technical usersAn AI assistant in the workspace that writes and runs code, builds pipelines and dashboards, and fixes errors
Genie OntologyWorks behind the othersA shared map of business meaning, such as what "active user" means, that Genie One and Genie Code both search
Genie ZeroOpsData engineersA background agent that watches jobs, pipelines and tables, finds the cause of a failure and proposes a fix. In Beta

How Genie answers a question

You type a question such as "what were sales by region last quarter?". Genie One finds a relevant Genie Agent or data asset, writes the SQL, runs it, and shows the result with the sources it used. You do not need to know SQL or how the workspace is organized.

Two things make the answer trustworthy, according to the docs:

  • It runs as you. The query uses your own permissions, so you only see data you are allowed to see. Row filters and column masks still apply.
  • It uses your company's definitions. Data teams set up Genie Agents with the right tables, example queries and business rules, and Genie Ontology supplies shared definitions. That is how "revenue" means the same thing in every answer.

Genie is therefore not a general chatbot. It is built to answer questions about your organization's data inside Databricks.

Genie One: for business users

Genie One is the simplified Databricks screen for people who do not work with notebooks, compute or queries. Besides chat, it shows the dashboards, Genie Agents, apps and tables shared with you in one place. It can also turn an answer into work: draft a document from a conversation, run a recurring report on a schedule, or take actions in connected tools such as Jira and Slack.

You can use it in the web browser, in the mobile apps for iOS and Android, in a desktop app, and in Slack. Our guide to the Genie One desktop app covers that version. Genie One was previously called Databricks One.

Genie Agents: where trusted answers are set up

A Genie Agent is a chat interface for one subject area, such as finance or sales. Data analysts build each one with datasets registered in Unity Catalog, example SQL queries, SQL expressions for business terms, and written instructions in the organization's own vocabulary. Users then ask questions and get back queries, result tables and charts.

Genie Agents were formerly called Genie Spaces, so older tutorials and exam guides may still use that name.

Genie Code: for developers

Genie Code is the AI coding and data assistant inside the Databricks workspace. It generates and runs code, builds pipelines and dashboards, debugs errors, and works with your Unity Catalog tables. It runs in notebooks, the SQL editor, the pipeline editor, dashboards and MLflow. The docs describe it as an autonomous agent: it plans, runs code, reads the output and fixes its own errors.

There is also a terminal version, Genie Code CLI, which we compare with the other Databricks command line tools in Agent Bricks CLI vs Genie Code CLI vs Unity Gateway CLI.

Is Databricks Genie free?

Partly, and the answer depends on the product:

  • Genie One and Genie Agents are free for users through January 31, 2027. Service principals, meaning automated accounts, are excluded and their usage is billed.
  • Genie Code has been billed pay-as-you-go since July 8, 2026. Each user gets a free amount of usage every month, and usage beyond it is charged.
  • Genie Ontology currently has no charge for curating its definitions, though the docs say that may change.

Databricks does not print the prices or the free amount in its docs; it points to its pricing page. Account admins can set budgets to track and control Genie Code spend.

Try it without a company account. Databricks Free Edition, the no-cost workspace for learners, includes Genie for dashboards and Genie Code for coding help. Our guide to learning Databricks for free explains how to sign up and what the limits are.

Why the name is confusing

Three things make "Genie" hard to search for. Databricks uses the word for the whole family, while many people use it for the chat feature alone. Two products were renamed: Databricks One became Genie One, and Genie Spaces became Genie Agents. And new members keep arriving, such as Genie ZeroOps. Our Genie One vs Genie Agents vs Genie Code comparison has the full timeline of names and a longer look at each product.

Common questions

Is Genie the same as Genie Code?

No. Genie Code is one member of the Genie family, made for developers. Business users asking data questions use Genie One.

Do I need to know SQL to use Genie?

No. Genie One writes and runs the SQL for you. Someone on the data team needs SQL skills to set up a Genie Agent well.

Can Genie see data I am not allowed to see?

No. The docs say queries run as you, so your Unity Catalog privileges, row filters and column masks apply.

Can other AI tools use Genie?

Yes. Agents built on other platforms can reach Genie through the Genie One MCP server, with the same permissions. See our article on the Genie One MCP server.

Sources

  1. Genie, Databricks docs, last updated November 18, 2026
  2. Genie One, Databricks docs, last updated November 18, 2026
  3. Genie Agents, Databricks docs, last updated November 19, 2026
  4. Genie Code, Databricks docs, last updated November 9, 2026
  5. Genie Ontology, Databricks docs, last updated November 16, 2026
  6. Genie ZeroOps, Databricks docs, last updated November 19, 2026
  7. Manage budgets and cost controls for Genie, Databricks docs, last updated November 16, 2026
  8. Use Genie on mobile, Databricks docs, last updated November 18, 2026

Checked against the Databricks documentation on November 21, 2026. Genie products are renamed and extended often, so verify details before you rely on them. brickster.ai is an independent community project and is not affiliated with Databricks.