Metabase is an open-source business intelligence platform that lets teams explore data, create visualizations, and build interactive dashboards without necessarily writing SQL. It connects to 30+ databases and provides a graphical query builder for non-technical users and a native SQL editor for analysts. Metabase has evolved significantly — with Data Studio for semantic-layer curation, Metabot AI for natural-language querying and SQL generation, a Modular Embedding SDK, and Actions for writing data back to databases — making it a full self-service analytics platform, not just a BI viewer. A key insight: models and transforms are Metabase's layered approach to reusable data: models add metadata within Metabase; transforms persist results as real database tables.
What This Cheat Sheet Covers
This topic spans 24 focused tables and 258 indexed concepts. Below is a complete table-by-table outline of this topic, spanning foundational concepts through advanced details.
Table 1: Core Concepts
Understanding Metabase's building blocks — from questions to documents to the semantic library — prevents confusion about where content lives and how the different content types relate to each other.
| Concept | Example | Description |
|---|---|---|
Query builder or SQL query | A query, its results, and its visualization — the basic analytical unit in Metabase. | |
Collection of charts with filters | Group related questions into tabbed pages with shared filters and interactive elements. | |
Turn question into a model | • Curated datasets built from questions • queryable like tables with annotated column metadata. | |
"Revenue" aggregation formula | • Standardized aggregations defined once and reused • ensures consistent calculations across teams. | |
"Marketing Analytics" folder | • Organizing folders for questions, dashboards, models, and metrics • permissions set at collection level. | |
Charts + narrative Markdown text | Reports that combine charts and rich text — like a Notion doc or Jupyter notebook — available on all plans. |