Metabase

Open source alternative to Tableau, Power BI, Looker, Sisense and ThoughtSpot

50.1K stars
Active this week
Last commit 1 day ago

Last verified Nov 3, 2026

Visualize, query, and share data with a self-hosted BI tool that lets your team explore databases without writing SQL.

Visit website GitHub50.1K Deploy on Hostinger

Read the full breakdownWho it is for, how it works, trade-offs, the quick start and the FAQ

who it's for

Who Metabase is for#

Data-curious business teams at growing companies

Metabase's question builder lets a marketer check conversion rates or a finance analyst review revenue trends without opening a SQL console. A single self-hosted instance can serve the whole company, and new users can be added with no licensing cost.

Skip if:

If your team already writes SQL fluently and needs notebook-style analysis with Python or R integration, a heavier tool like Apache Superset may fit better. Metabase's strength is accessibility, not data-science depth.

SaaS companies building embedded customer analytics

The React SDK embeds dashboards, charts, and a question builder directly into your product under your own branding. Multi-tenant JWT auth ensures each customer sees only their own rows. This saves the engineering work of building an analytics layer from scratch.

Skip if:

If your product needs highly custom visualization types or real-time streaming data, Metabase's embedded analytics may not cover the edge cases. It is strongest for SQL-backed datasets with standard chart types.

Data teams setting up a governed semantic layer

Data Studio lets senior analysts define the canonical metrics and business logic that every dashboard and AI query must follow. Git sync versions those definitions. This is useful when different teams are building dashboards and you need consistency across reports at scale.

Skip if:

If your team has fewer than five people and dashboards are built by one or two analysts, the overhead of maintaining a formal semantic layer in Data Studio is probably not worth it at this stage.

Teams giving AI agents access to company data

The MCP server exposes Metabase's query and dashboard capabilities to Claude Code, Codex, and Cursor. An agent can read published metrics, build dashboards, and answer analytical questions while staying within each user's row-level permissions. No separate API key or separate data exposure is required.

Skip if:

If your data team has not yet built out the semantic layer and canonical metrics, an AI agent querying raw tables will produce inconsistent answers. The MCP is most valuable after the semantic layer is in place.

the problem

The problem it solves#

Most teams with data questions still depend on a single analyst or engineer to write SQL queries for them. The wait for a report can stretch hours or days, and the analyst becomes a bottleneck between the data warehouse and the business decision that depends on it.

The underlying frustration is that the data is there; the access layer is not. A finance manager cannot explore quarterly trends without filing a ticket. A product manager cannot compare cohort behavior without asking an engineer. Business intelligence tools from vendors like Tableau and Power BI solve the access problem, but they carry per-seat licensing costs that price out every casual user and lock the data team into a managed cloud.

how Metabase solves it

How it solves it#

No-SQL question builder

Lets anyone on the team query a connected database by clicking through a visual interface instead of writing SQL. Fields, filters, and aggregations are point-and-click, so a marketing analyst can answer "how many new signups came from organic last week?" without filing a request to the data team.

Metabot AI assistant

Answers questions in plain English against your actual data, guided by the metrics and business logic you define in Data Studio. Available on every plan, including the AGPL community build, with a choice of AI provider or a self-hosted model. Works in Metabase, via the MCP server, or through the CLI.

Semantic layer in Data Studio

Your data team defines canonical tables, metrics, and business logic once in Metabase's Data Studio. Every dashboard, Metabot answer, and agent query builds on those verified definitions. Git sync keeps the semantic layer version-controlled so changes go through a review process.

Embedded analytics SDK

Publish dashboards, charts, or a full self-serve analytics experience inside your SaaS product using the React SDK. Each embedded session is scoped to that customer's data via multi-tenant JWT or SAML auth, and the whole surface is white-labeled to your brand.

MCP server and CLI for AI agents

Exposes your data to Claude, Codex, Cursor, or any MCP-compatible coding agent through the Metabase MCP server. The agent can query metrics, build dashboards, and write data apps while staying within each user's row-level permissions, so AI access inherits your existing governance.

Alerts and dashboard subscriptions

Set threshold alerts that trigger when a query crosses a value, or schedule dashboards to deliver to email, Slack, or a webhook on a fixed cadence. Alerts and subscriptions work on the community edition without additional infrastructure or a paid plan.

strengths · trade-offs

Strengths and trade-offs#

Strengths

No per-seat licensing on the AGPL editionThe community build (Docker Hub `metabase/metabase`) is AGPL-3.0 and free to run on your own server. Every person on your team can log in and explore data without adding to a license bill. Commercial BI tools like Tableau and Looker charge monthly per seat, which makes broad access expensive at any team size.
Five-minute setup with DockerThe README describes a five-minute setup: one Docker command, a database to connect, and the web UI opens at localhost:3000. No agents, no cloud configuration, and no professional services requirement for a basic deployment. The maker documents the same path as the primary install.
50,000+ stars and over 10 years of active developmentThe GitHub repository has over 50,000 stars and was created in February 2015, with a commit as recent as November 4, 2026. That history means a large corpus of community documentation, answered issues, and third-party integrations covering most common databases and deployment patterns.
AI access with permission guardrailsMetabot and the MCP server give AI tools and agents governed access to data, scoped to each user's existing permissions. A coding agent working through the MCP server cannot read rows the human user cannot see, which removes a common security objection to giving AI tools direct database access.

Trade-offs

-Split license: enterprise features require a commercial licenseThe code outside the top-level `enterprise/` directory is AGPL-3.0, freely self-hostable for internal commercial use at no license cost. The enterprise directory and the enterprise Docker image are under the Metabase Commercial License, which restricts reselling Metabase as a managed/hosted service. Features like SAML SSO, row-level permissions, and audit logs sit behind the enterprise tier.
-High open issue countThe repository carries 3,985 open issues. For a project of this scale that is not unusual, but it means some edge cases in connectors, embeddings, and the query builder may have known bugs without a near-term fix date. Teams relying on specific database drivers should check the issue tracker before committing.
-Resource requirements for production self-hostingA production Metabase instance needs a dedicated application database (Postgres is recommended), sufficient RAM for the JVM, and a reverse proxy for TLS. The Docker quick-start is enough for evaluation but not recommended as a production configuration; plan for a proper deployment before rolling it out to your whole team.

versus alternatives

Metabase vs alternatives#

Metabase vs Tableau

Both tools serve non-technical users querying SQL databases, but they target different team profiles. Metabase is built for self-hosted, broad team access with a short setup time; Tableau targets analysts who need advanced statistical visualization and a large library of certified connectors.

Feature	Metabase	Tableau
License	AGPL-3.0 (community)	Proprietary
Self-hosting	Yes	No
Pricing	Free community; paid Pro/Enterprise	Per-seat subscription
No-SQL builder	Yes	Yes (drag-and-drop)
Embedded analytics	Yes (React SDK)	Yes (Tableau Embedded)

Metabase is the stronger choice when cost or data-residency requirements rule out a per-seat SaaS subscription and when most users are non-technical. Tableau is worth considering when analysts need advanced statistical functions, R or Python integration, or access to Tableau's certified connector ecosystem.

Metabase vs Power BI

Power BI integrates natively with Microsoft 365, SharePoint, and Azure data services. Teams already standardized on the Microsoft stack often reach for it because of those native connectors.

Feature	Metabase	Power BI
License	AGPL-3.0 (community)	Proprietary
Self-hosting	Yes	No (Report Server is limited)
Pricing	Free community; paid plans	Per-user licensing
Native Microsoft connectors	No	Yes

Metabase is the better fit for teams outside the Microsoft ecosystem who want a fully self-hosted deployment with no per-user fees. Power BI is the stronger choice when the organization is standardized on Microsoft 365 and analysts already work in Excel.

Metabase vs Looker

Looker (Google Looker) is a semantic-layer-first BI tool that runs fully in the cloud and uses LookML, a proprietary data modeling language, to enforce consistency across reports.

Feature	Metabase	Looker
License	AGPL-3.0 (community)	Proprietary
Self-hosting	Yes	No
Semantic layer	Data Studio (built-in)	LookML (requires modeling expertise)
Pricing	Free community; paid plans	Enterprise pricing

Metabase is a practical choice for mid-sized teams that want a governed semantic layer without LookML expertise or Looker's enterprise pricing. Looker is worth considering when the team needs deep LookML modeling or is already embedded in Google Cloud infrastructure.

install · quick start

Quick start#

bash

Self-hosting uses Docker with the official AGPL community image from Docker Hub.

```bash
docker run -d -p 3000:3000 --name metabase metabase/metabase
```

To deploy without Docker, download `metabase.jar` and start with the Java runtime:

```bash
java --add-opens java.base/java.nio=ALL-UNNAMED -jar metabase.jar
```

tech stack · detected from GitHub

What it's built on#

Languages: JavaScriptTypeScript
Frameworks: ExpressNext.jsReact
Databases: MySQLPostgreSQL
Tooling: esbuildWebpack

frequently asked

FAQ#

Is Metabase free to use?

The community edition is AGPL-3.0 and free to self-host at no licensing cost. Everything outside the top-level enterprise/ directory in the repository is covered by the AGPL, and the Docker Hub image (metabase/metabase) is the AGPL build. Pro and Enterprise plans add features like SAML, row-level permissions, and audit logs, with pricing listed on Metabase's site.

What databases does Metabase connect to?

Metabase officially supports PostgreSQL, MySQL, MariaDB, SQL Server, BigQuery, Snowflake, Redshift, Athena, and others; a full list is in the repository under docs/databases/connecting.md. Community-built drivers extend coverage to additional databases. Each database needs only a read-access connection string; Metabase queries it at runtime and does not store a copy of your data.

Can I embed Metabase in my own product?

Yes. The React embedding SDK (@metabase/embedding-sdk-react) lets you drop dashboards and charts into your app. Multi-tenant JWT or SAML auth scopes each session to that customer's own data. Full white-labeling, custom fonts and colors, and row-level permissions are available. Production embedding at scale uses the enterprise Docker image, which is under the Metabase Commercial License.

How is Metabase different from Tableau or Power BI?

Tableau and Power BI are proprietary tools with per-seat licensing and vendor-managed hosting. Metabase's community edition is AGPL-3.0, runs on your own server, and has no per-seat fees. The tradeoff is that Tableau and Power BI have more advanced statistical modeling and larger ecosystems of certified connectors. Metabase wins on ease of setup, no-SQL accessibility, and cost for broad team access.

What does the AGPL license mean for self-hosting?

You can run the AGPL community edition on your own infrastructure, internally, for any commercial purpose, with no license fee. The AGPL restriction applies if you distribute a modified version externally as a service: you must release your modifications under the AGPL. Reselling Metabase itself as a hosted/managed service requires a commercial license. The enterprise directory and enterprise Docker image are under the Metabase Commercial License, not the AGPL.

also worth a look

Similar open-source tools#

Redash

Alternative toTableau

Connect to any data source, query, visualize, and share

28,864 GitHub stars
Licence BSD-2-Clause

Apache Superset

Alternative toMicrosoft Power BI

Open source data visualization and BI exploration tool

75,404 GitHub stars
Licence Apache-2.0

MindsDB

Alternative toLindy AI

Delegate reports, inbox cleanup, and workflow tasks to an AI agent

758 GitHub stars
Licence MIT

Lightdash

Alternative toLooker

Self-serve BI for dbt projects with metric definitions

6,221 GitHub stars
Licence MIT

Cube

Alternative toAtScale

Headless BI platform with a shared semantic layer for your data

21,143 GitHub stars

MiroFish

InAI & Machine Learning

Multi-agent simulation engine that predicts real-world outcomes

83,421 GitHub stars
Licence AGPL-3.0

by category

Open source tools in these categories#

113

Self-host Metabase on a Hostinger VPS

Keep it running 24/7, even when your laptop is off. A Hostinger VPS gives it a server of its own.

01
Choose a plan
Starts from $6.49/mo
02
Apply
Extra 10% off yearly plans
03
Install Metabase
On a server that stays on
04
Start building
Focus on the app, not the server

Deploy on Hostinger

Inbox Zero

Alternative toSuperhuman

AI email assistant that drafts replies and clears your inbox

12,889 GitHub stars
Licence AGPL-3.0
#Email

rakazo

Alternative toGrok Bot

AI teammates you own: your keys, your model, your machine.

6,568 GitHub stars
Licence Apache-2.0
#AI Agents

SnitchFeed

Live monitoring on LinkedIn, X, Reddit & more. Agentic social listening for startups.

Marketing & Growth

Advertise your product

The projects we add and every new guide. Free, unsubscribe any time.

Find open source alternatives to paid software

Metabase vs Tableau

Feature	Metabase	Tableau
License	AGPL-3.0 (community)	Proprietary
Self-hosting	Yes	No
Pricing	Free community; paid Pro/Enterprise	Per-seat subscription
No-SQL builder	Yes	Yes (drag-and-drop)
Embedded analytics	Yes (React SDK)	Yes (Tableau Embedded)

Metabase vs Power BI

Power BI integrates natively with Microsoft 365, SharePoint, and Azure data services. Teams already standardized on the Microsoft stack often reach for it because of those native connectors.

Feature	Metabase	Power BI
License	AGPL-3.0 (community)	Proprietary
Self-hosting	Yes	No (Report Server is limited)
Pricing	Free community; paid plans	Per-user licensing
Native Microsoft connectors	No	Yes

Metabase vs Looker

Looker (Google Looker) is a semantic-layer-first BI tool that runs fully in the cloud and uses LookML, a proprietary data modeling language, to enforce consistency across reports.

Feature	Metabase	Looker
License	AGPL-3.0 (community)	Proprietary
Self-hosting	Yes	No
Semantic layer	Data Studio (built-in)	LookML (requires modeling expertise)
Pricing	Free community; paid plans	Enterprise pricing

Who Metabase is for#

Data-curious business teams at growing companies

Skip if:

SaaS companies building embedded customer analytics

Skip if:

Data teams setting up a governed semantic layer

Skip if:

Teams giving AI agents access to company data

Skip if:

The problem it solves#

How it solves it#

No-SQL question builder

Metabot AI assistant

Semantic layer in Data Studio

Embedded analytics SDK

MCP server and CLI for AI agents

Alerts and dashboard subscriptions

Strengths and trade-offs#

Strengths

Trade-offs

Metabase vs alternatives#

Metabase vs Tableau

Metabase vs Power BI

Metabase vs Looker

Quick start#

What it's built on#

FAQ#

Similar open-source tools#

Open source tools in these categories#

Data & Analytics

Business & Productivity

AI & Machine Learning

Find open source alternatives to paid software

Who Metabase is for#

Data-curious business teams at growing companies

Skip if:

SaaS companies building embedded customer analytics

Skip if:

Data teams setting up a governed semantic layer

Skip if:

Teams giving AI agents access to company data

Skip if:

The problem it solves#

How it solves it#

No-SQL question builder

Metabot AI assistant

Semantic layer in Data Studio

Embedded analytics SDK

MCP server and CLI for AI agents

Alerts and dashboard subscriptions

Strengths and trade-offs#

Strengths

Trade-offs

Metabase vs alternatives#

Metabase vs Tableau

Metabase vs Power BI

Metabase vs Looker

Quick start#

What it's built on#

FAQ#

Similar open-source tools#

Open source tools in these categories#

Data & Analytics

Business & Productivity

AI & Machine Learning

Find open source alternatives to paid software