Self-host Tableau

Business intelligence / dashboards / data exploration · Category: storage, queues & search

Tableau is the long-standing leader in BI dashboards, with a strong drag-and-drop authoring flow and broad data-source coverage. The self-hostable replacements all cover the SQL-against-warehouse path well; the gap is mostly polish, governance, and the (frankly extreme) Tableau Desktop authoring fluency.

Tableau pricing anchor: $15-75/user/mo on Tableau Cloud — Creator seats hit ~$75/mo and stack up fast across an org.

Metabase metabase/metabase alive

GitHub: ★ 47.3k · last commit 1d ago · 4157 open issues
License: AGPL-3.0
Core is AGPL-3.0; some enterprise features (sandboxing, advanced permissions, white-labeling) are commercial-only on the Pro/Enterprise editions.
Setup time: 10min — single JAR or docker-compose
Monthly cost: $10 VPS handles a small team; Postgres app DB is light. Heavier queries cost on the warehouse, not Metabase.

Migration sketch. Tableau workbooks (.twbx) don't import. Practical path: list your Tableau dashboards by importance; in Metabase, connect the same data warehouse (Snowflake / Postgres / BigQuery / etc.), recreate top dashboards using Metabase's question builder or native SQL. Question→dashboard composition is similar to Tableau worksheets→dashboards.

Good fit forTeams that want most analysts to self-serve via a friendly UI; non-technical users get a gentler ramp than Tableau Desktop.

Weak atTableau-grade visualizations (advanced chart calc, level-of-detail expressions, complex drill-throughs) are not all there.

Apache Superset apache/superset alive

GitHub: ★ 72.8k · last commit today · 1289 open issues
License: Apache-2.0
Setup time: 30min docker-compose (Superset + Postgres + Redis + Celery worker)
Monthly cost: $15 VPS for the multi-container stack; production setups want a dedicated Postgres + Redis.

Migration sketch. Superset reads from the warehouse like Tableau does. Connect data sources via SQLAlchemy URIs; create datasets (semantic layer); recreate dashboards via the chart editor. There is no .twbx importer. The DAX-equivalent layer (Superset metrics + virtual datasets) approximates Tableau calculated fields.

Good fit forEngineering-heavy teams that want SQL-first authoring, Jinja-templated queries, and embedding into other apps via dashboards-as-iframe.

Weak atDrag-and-drop chart authoring is less fluid than Metabase or Tableau; permissions model is finicky.

Redash getredash/redash alive

GitHub: ★ 28.6k · last commit 3d ago · 753 open issues
License: BSD-2-Clause
Setup time: 20min docker-compose (Redash server + Postgres + Redis + worker)
Monthly cost: $10 VPS; lighter than Superset but heavier than Metabase.

Migration sketch. Redash is query-first: write SQL, save, visualize, drop on a dashboard. Tableau workbook authors who already lean on raw SQL transition cleanly. Connect the warehouse, recreate top dashboards by porting the SQL behind each Tableau view (the calc engine is in your warehouse, not in Tableau or Redash).

Good fit forTeams whose Tableau use is essentially canned SQL → chart, where Tableau Desktop's authoring fluency was overkill.

Weak atProject pace has slowed since Databricks acquired the original team; check the freshness pill before adopting.

In a terminal? npx os-alt tableau prints this table — how the CLI works →