Metabase
Metabase is open source business intelligence and embedded analytics. For the open source or self-hosted route, we hold it: Grafana is the better default to limit vendor lock-in on observability and lighter analytics. Metabase Cloud SaaS can still fit when data is spread across many systems and a managed BI layer beats running another platform yourself. Do not start new OSS Metabase deployments when Grafana already serves the team.
Blurb
Metabase is the easy, open-source way for everyone in your company to ask questions and learn from data.
Summary
What it is: Web app plus Metabase Cloud (SaaS). Visual query builder, native SQL editor, dashboards, pulses, and embedding APIs. Open Source (AGPL) and commercial editions share one codebase.
When to use (SaaS): Data spread across many databases or clouds; want managed hosting, upgrades, and support without operating BI infra; embedded analytics for customers with a vendor-run control plane.
When to skip (OSS): Greenfield self-hosted BI when Grafana (trial) already covers metrics and enough SQL-backed panels. Prefer Apache Superset for large semantic layers or Redash for engineer-first SQL sharing.
Hold (OSS): Existing self-hosted Metabase is acceptable. New open source rollouts should default to Grafana unless requirements are explicitly warehouse-centric BI that Grafana cannot cover.
Key features: No-code questions, drill-through, collections and permissions, optional AI assistant (Metabot), Docker or JAR deploy in minutes.
Details
| Topic | Notes |
|---|---|
| Deploy | Hold path: Docker/JAR self-hosted (prefer Grafana OSS instead for new work). SaaS path: Metabase Cloud when federated data or ops cost dominates |
| Data model | Live queries against connected databases; extracts optional via caching settings |
| Auth | Google, LDAP, SAML; row-level sandboxing on supported plans |
Practices: Use read-only DB credentials; separate analytics replicas from OLTP; audit public links and embedded iframe origins.
References