Databricks Genie One: pricing, connectors and setup

Genie One is the screen Databricks gives business users: a chat box that answers data questions in plain words, with dashboards, apps and shared documents beside it. It used to be called Databricks One. Until 31 January 2027 it is free for every user; you pay only for the SQL warehouse that runs the queries behind the answers, and for any use by service principals.

It is one of 3 Genies. Genie vs Genie One vs Genie Code explains the family; this page is about the one most people in a company will open.

What Genie One does

• Answers questions. Chat looks for a Genie Agent that fits the question first, then searches dashboards, saved queries and metric viewsSemantic LayerOne shared definition of business terms and metrics, such as revenue or active customer, that reports, dashboards and AI assistants all read instead of writing their own.. Every answer cites what it used.
• Opens what is shared. Dashboards, Databricks Apps and Genie Agents open inside Genie One, read only unless you can edit them.
• Reads your files and tools. Upload CSV, Excel, PDF, Word, PowerPoint or images to a conversation, or connect Google Drive, Microsoft 365, Slack, Atlassian, GitHub and others.
• Works on a schedule. "Send me a daily briefing of new customer reviews" becomes a scheduled task that posts to a thread and emails the result with a PDF.
• Writes things up. It drafts a document from a conversation, which you edit and share, and it can run code in an isolated sandbox for analysis that SQL cannot do.
• Remembers. Ask it to remember that your fiscal year starts in February and it applies that to later chats. Memories are private to you.

What it costs

Databricks says Genie One has no per-seat price. To see how much free use people make of it, query the billing table:

SELECT usage_date, identity_metadata.run_as AS user_name, SUM(usage_quantity) AS dbus
FROM system.billing.usage
WHERE billing_origin_product = 'GENIE'
  AND sku_name = 'GENIE_FREE_USAGE'
  AND usage_metadata.genie.surface = 'GENIE_ONE'
GROUP BY ALL
ORDER BY usage_date DESC;

Free rows only start on 20 July 2026. This is worth running before February 2027, so you know what the use would cost when the free period ends. The Genie Code side, with its own budgets, is in Genie Code pricing.

Who needs which access

Consumer access is the smallest entitlement, so you can give Genie One to a whole company without opening notebooks or clusters to anyone. Answers follow Unity Catalog permissions, so a question about a table someone cannot read does not get an answer from it.

Try which Genie a task needs

Try itWhich Databricks Genie do you need?

The job

·Ask how sales did last week·Find the right dashboard·Answer from company documents·Set up trusted answers for a team·Define revenue once for everyone·Build a pipeline·Fix a failing notebook

Your access

Consumer accessWorkspace or SQL access

1. Genie One

   Use this

   Business users

   Its own simple UI at /one: chat, dashboards, apps, mobile

   Free for users until 31 Jan 2027; the SQL warehouse is billed

2. Genie Agents

   Data teams who set up trusted answers

   A chat over chosen tables, with example SQL and instructions

   Free for users until 31 Jan 2027; the SQL warehouse is billed

3. Genie Code

   Engineers, analysts and data scientists

   Notebooks, SQL editor, pipelines editor, dashboards, MLflow

   150 free DBUs a user a month, then pay as you go

Chat in Genie One looks for a Genie Agent that fits the question first, then searches dashboards, queries and metric views.

Set it up in 5 steps

1. Give people Consumer access in each workspace, and CAN USE on a serverless SQL warehouse. Chat picks a warehouse itself unless the user chooses one.
2. Build a few Genie Agents first. Chat goes to an agent before anything else, and a curated agent with example SQL answers far more reliably than a search across every table. Keep them few: Databricks notes that a workspace with many agents can route questions less accurately.
3. Define your numbers once as metric views, so "revenue" means the same in chat, dashboards and Genie Code. Databricks metric views shows how.
4. Write the house rules in /Workspace/.genie_workspace_instructions.md, under 20,000 characters: your terms, fiscal calendar and how answers should look. It applies to chat only, not to agents or Genie Code.
5. Decide on account level. Account-level Genie One, at accounts.azuredatabricks.net/one on Azure, shows everything a person can open across workspaces. It runs on a managed workspace named databricks-system-workspace, and some Databricks metadata about it can be processed in the US. If your rules forbid that, an account admin turns it off under Settings, Feature enablement.

Connectors to outside tools sign in as each user, never with a shared token, and every tool call goes through Unity Gateway, which can require approval and logs it.

Where its answers come from

Genie One and Genie Code read the same store of business context, the Genie Ontology: metric views and domains you define, plus snippets Genie learns from your dashboards, queries and agents. That is why the cheapest way to better answers is better definitions, not a better prompt.

If you are building the data underneath, our data engineering course has a part on dashboards and Genie over a real gold layer.