Metabase Pro: Features, Costs, and When You Need It

Metabase Pro unlocks row-level security, embedded dashboards, and white-labeling for $575 per month (10 users, then $12 per additional user). It sits between the free Open Source tier and Enterprise, targeting teams that need to control who sees what data or embed analytics into customer-facing products. If your team is 20+ people or you sell analytics to customers, Pro is probably your tier.

What is Metabase Pro?

Metabase Pro is the standard paid tier of Metabase, a self-hosted or cloud-hosted business intelligence tool. Metabase is used by over 100,000 companies to build dashboards and reports without writing SQL. Pro adds three capabilities the free tier blocks: row-level security (so each user sees only their data), interactive embedding (for SaaS companies building analytics into their product), and white-labeling (removing Metabase branding). The trade-off is cost: $575/month billed monthly, or $517.50/month billed annually.

Metabase itself is not fancy. It does not do machine learning. It does not do predictive analytics. It builds dashboards, charts, and reports from databases or APIs. Its value is speed: your team can ship a dashboard in hours, not weeks. Pro adds governance and distribution.

When You NEED Metabase Pro: The Three Feature Gates

Pro unlocks three things. Open Source and Starter (a lower-priced tier at $100/month) do not have them. The jump from Starter ($100) to Pro ($575) is $475/month. You only make that jump if you actually need at least one of these.

1. Row-Level Security (RLS): Your users log into Metabase. RLS means Alice sees only her region's sales, Bob sees only his customers' data. The database still has all the data; Metabase filters it per user. Open Source has no RLS. Starter has basic permissions (can this person view this dashboard?) but not row-level filtering. Enterprise has it. Pro has it. If you have 50 internal users and each needs a different slice of the same dashboard, you need Pro for RLS.

2. Embedding and Interactive Dashboards: You embed Metabase dashboards into your SaaS product or intranet. Your customers log into your app, click "Reports," and see their Metabase dashboard without leaving your app. Interactive embedding means they can filter, drill down, and export right there. Open Source has basic, read-only embedding. Pro adds interactive embedding. Enterprise adds more whitelisting controls. If you sell B2B SaaS and your customers expect to see their own analytics, you need Pro embedding.

3. White-Labeling: Remove all Metabase branding. The interface does not say "Powered by Metabase." The logo is yours. The colors are yours. Open Source and Starter always show Metabase branding. Pro removes it. Enterprise adds more customization. If your customers will see the interface, white-labeling matters. If it's internal only, it does not.

If you do not need any of these three, the free Open Source tier might be enough. If you need one or two, Pro is the right level. If you need everything plus custom SAML, advanced support, and audit logs, you are in Enterprise territory (around $20,000-$39,000 per year).

Permissions and Security: Row- and Column-Level Data Access

Metabase Pro's permissions model is table-stakes, not best-in-class. Understand what it does and does not do.

Row-level filtering works like this: your database has a sales table with columns region, amount, date. You define a rule: "Alice sees only rows where region = 'North."' Alice logs into Metabase, runs any query or views any dashboard using that table, and only North data comes back. The rule is centralized in Metabase; you do not rewrite queries.

Column-level filtering hides columns. Alice's sales table does not show the price_cost column, only price_retail. Bob sees both.

Group-based access: create a group called "North Sales" and assign 15 people to it. Assign a row-level rule to the group. All 15 inherit it. Update it in one place.

What Pro does NOT do: it does not handle complex, context-dependent rules. "Alice sees only rows where region = her profile's region" requires a database-backed attribute. Metabase can do it, but setup is manual. It does not scale to hundreds of fine-grained, user-specific rules across dozens of tables. For that, you need a dedicated fine-grained access control platform or database-native solutions like row-level security in PostgreSQL or column-level masking in Snowflake.

Pro also includes audit logs (who ran what query, when) and password policies. Enterprise adds SSO and advanced compliance features.

Opsily's managed Metabase hosting handles the infrastructure side so you focus on the access control logic itself.

Embedded Analytics: Building Customer-Facing Dashboards

If you build SaaS, embedded analytics is a serious feature. Your customers do not log into your Metabase instance. They log into your product. Behind the scenes, your app makes an API call to Metabase to fetch a pre-built dashboard or chart, then renders it in an iframe or embeds it natively.

Open Source can embed read-only dashboards. Starter adds a bit more. Pro adds interactive embedding: your users can click filters, drill into data, and export without leaving your app. They cannot see other customers' data; Metabase enforces that via row-level security.

White-labeling is critical here. If your customer sees a Metabase logo or watermark, they know you are using Metabase. That is fine if you are transparent about it. But if you want analytics to feel like your product, Pro removes the branding.

Two common patterns:

Pattern 1: SaaS Dashboard. You are a project management SaaS. Each customer account has a dashboard showing their project spend, team velocity, and risk flags. You embed a Metabase dashboard, filtered to that customer's data via session variables or API parameters. Your customer sees a dashboard that looks like part of your app. They do not know it is Metabase.

Pattern 2: Internal Analytics Portal. You have 80 employees. You build a Metabase dashboard showing daily revenue, churn, CAC by cohort. You embed it in your intranet or a dedicated analytics page. RLS ensures each region manager sees only their region's KPIs.

Both need Pro for interactive features. Read-only embedding does not feel like part of your product; it feels like a separate tool.

Metabase's API documentation covers session tokens and embedding parameters in detail.

Total Cost of Ownership: License Plus Infrastructure

Metabase Pro costs $575/month. That is not the full cost of running it.

If you use Metabase Cloud (Metabase's hosted service), the $575 includes hosting. You pay a fixed amount, and Metabase handles servers, backups, and scaling. Simple. No infrastructure bill.

If you self-host Metabase Pro (more common for teams with compliance or latency needs), you pay for:

1. Metabase Pro license: $575/month (or $517.50/month annual)
2. Compute: A server or container to run Metabase itself. Minimum viable: t3.small on AWS ($20/month) or a $5/month DigitalOcean droplet. Realistic production: t3.medium ($40/month) or a 2-CPU Docker container.
3. Database: The database Metabase connects to (PostgreSQL, Snowflake, BigQuery, etc.). That cost is separate from Metabase.
4. Storage: Metabase internal database (usually PostgreSQL) to store dashboards, queries, users. ~50 GB typical; negligible on most plans.
5. Backups: Automated snapshots of Metabase's metadata. ~$5-$20/month on AWS or similar.
6. Operations: A junior engineer or DevOps person spending 2-4 hours per month on upgrades, backups, and monitoring. At $100/hour, that is $200-$400/month.

Full self-hosted monthly cost: $575 + $40 (compute) + $10 (storage) + $300 (ops labor) = ~$925/month. A small team might get it lower ($575 + $40 + $0 ops = $615 if someone learns it gradually). A large team with strict SLAs might pay more.

Cloud cost: $575/month, flat. You trade operational overhead for loss of control. No custom configurations. Slower response to scaling issues. But no on-call responsibilities.

Compare this to alternatives:

• Tableau: $70-$120/user/month. A 20-person analytics team: $1,400-$2,400/month, minimum. No setup, heavy licensing.
• Power BI: $10-$25/user/month, plus Azure costs. Smaller per-user bill, but still scales with headcount.
• Looker (Google Data Studio enterprise): ~$500-$2,000/month for basic usage, plus GCP compute.

Metabase Pro's cost advantage is flat-rate: you pay $575 no matter if you have 5 or 50 users (as long as you have the Pro license for 10-user base). You only pay extra if you add users beyond 10 at $12/user/month.

Pro vs. Starter vs. Enterprise: Which Tier Makes Sense

Metabase has four tiers: Open Source (free, self-hosted only), Starter ($100/month), Pro ($575/month), and Enterprise (custom pricing, typically $20,000-$39,000 per year).

The most common jump is Starter Pro. You hit it when either (a) you have too many users for simple "all or nothing" permissions, or (b) you want to embed analytics. $475 extra per month is steep if you do not need it. Painful but survivable if you do.

The Pro Enterprise jump is rare. You make it because your security/compliance team requires SAML, advanced audit logs, or your support contract is mission-critical.

Getting Started: Deployment Options (Cloud vs. Self-Hosted)

Once you buy a Pro license, where do you run it?

Metabase Cloud: Hosted by Metabase, available in US, Europe, Latin America, and Asia-Pacific regions. You create an account, connect your database, and start building dashboards. Metabase handles patching, backups, and scaling. Trade-off: zero control over infrastructure, slower to customize, but zero ops work.

Self-Hosted (Docker): You run Metabase on your own servers or Kubernetes. Full control. Faster to debug. Meets compliance requirements (data stays on your network). Trade-off: you manage updates, backups, and scaling. Most common deployment uses Docker, Docker Compose or Kubernetes for orchestration.

Self-hosted is more common for teams that (a) have strict data residency rules, (b) want to version-control their Metabase setup, (c) run Kubernetes and want tight integration, or (d) have highly custom database setups and need to debug networking.

Cloud is simpler if you have no special requirements and want to reduce ops overhead.

Setup steps for self-hosted:

1. Provision a small VM or container (t3.small or 1-CPU container).
2. Install Docker.
3. Write a docker-compose.yml that specifies the Metabase image and a PostgreSQL database for Metabase's metadata.
4. Run docker-compose up.
5. Navigate to the Metabase interface (on localhost:3000), create an admin account, and license it with your Pro key.
6. Connect your data warehouse (BigQuery, Snowflake, PostgreSQL, etc.).
7. Start building dashboards.

Time investment: 1-2 hours for a first-timer. Days if you have network policies or custom authentication.

Real-World Use Cases

Case 1: B2B SaaS with Embedded Analytics. A project management SaaS with 300 customers. Each customer wants to see their project spend, team utilization, and timeline risk. The product team builds a Metabase dashboard, white-labeled to match the app's UI, and embeds it per-customer using RLS. The dashboard shows only that customer's projects. Cost justified: Pro ($575/month) + self-hosted infrastructure ($50/month) = $625/month. Without it, they would need a custom analytics microservice (engineer-months of work).

Case 2: Mid-Market Internal Analytics. A 100-person SaaS company has sales, product, finance, and ops teams. Each team needs different slices of the same revenue and product-usage data. Sales wants to see by region and customer. Finance wants to see by quarter and contract value. Product wants to see by feature and cohort. With Starter, they would need 4 separate dashboards. With Pro and RLS, they build one source-of-truth dashboard and Metabase filters it automatically per team. Cost justified: Pro ($575/month) feels cheap compared to hiring a BI analyst ($150K+ per year) to build and maintain separate reports.

Frequently Asked Questions

How much is Metabase Pro?

Metabase Pro costs $575 per month billed monthly, or $517.50 per month billed annually (10% discount). This covers 10 named users. Each additional user costs $12 per month. A 30-person team would cost $575 + (20 x $12) = $815/month.

Is Metabase free or paid?

Metabase is both. The Open Source tier is free and open-source; you can self-host it forever without paying. Starter is $100/month. Pro is $575/month. Enterprise is custom. Most teams start with free, then move to Starter or Pro as they grow.

What are the drawbacks of Metabase?

Metabase is not a replacement for a data warehouse. It is a query builder and dashboarding tool. It is slower than dedicated analytics databases on large datasets. It does not do machine learning or anomaly detection. Its permission model is simpler than enterprise BI tools. It does not have a mobile app. If you need bleeding-edge features, you will hit limits.

How do I deploy Metabase Pro self-hosted?

Docker is the standard. Write a docker-compose.yml with Metabase and PostgreSQL, run docker-compose up, then navigate to port 3000 and set up your admin account and license. Kubernetes is also supported. See the official docs for detailed instructions.

Can I embed Metabase dashboards without white-labeling?

Yes, but your customers will see the Metabase logo. White-labeling is a Pro feature that removes branding. If branding does not matter, Starter might be enough.

Does Metabase connect to my data warehouse?

Yes. Metabase connects to PostgreSQL, MySQL, MongoDB, Snowflake, BigQuery, Redshift, Athena, and dozens of other databases and APIs. You point Metabase to your database, it pulls metadata, and you start building dashboards.

What databases does Pro support?

Pro supports the same database integrations as Open Source and Starter. The tier difference is not database support; it is permissions, embedding, and white-labeling.

The Bottom Line

Metabase Pro costs $575 per month and adds row-level security, embedded dashboards, and white-labeling. You need it if you have multiple internal teams that need different data views, or if you sell embedded analytics to customers. If you are building a simple internal dashboard for a small team, the free tier might be enough. If you need advanced governance or SAML, you are in Enterprise territory.

The real cost is not just the license: it is infrastructure, operations, and time. Self-hosting saves money but requires DevOps work. Cloud hosting removes that burden but costs more and gives you less control.

Start with the Opsily managed Metabase hosting guide to understand infrastructure requirements and deployment options for your team size.