Power BI supported data sources and how to pick the right one
Power BI supported data sources are the files, databases and online services that Power BI can read through its built-in connectors. The list runs into the hundreds, from Excel and SQL Server to Salesforce and Fabric lakehouses. But support is not all the same. Some sources allow live queries, some need a gateway, and some refresh only in Desktop. So a quick check before you build saves rework later.
- Connector groups at a glance
- Modes and gateways
- Checks before you build
Last updated
What counts as Power BI supported data sources?
Power BI reads data through Power Query connectors. Each connector knows how to sign in to one kind of source and turn its data into tables. In Desktop, you see them under Get data, grouped by type. Dataflows in the service use the same engine, although their list differs a little.
So a source is supported when a connector exists for it and Microsoft documents it. Some connectors come from Microsoft, while others come from the vendor of the source. A few carry a Beta or Preview label, which means they may change before general release.
When no connector exists, you still have routes. For example, generic connectors such as ODBC, OData and Web reach many systems. Otherwise, a staging database or a file export can sit between the source and Power BI.

Which Power BI supported data sources fall in each group?
Get data sorts connectors into groups. These examples show the usual members, not the full list.
| Group | Common examples | Good to know |
|---|---|---|
| File | Excel, CSV, XML, JSON, PDF, folder | Files on a laptop need a gateway to refresh |
| Database | SQL Server, Oracle, PostgreSQL, MySQL, Snowflake | Many allow DirectQuery as well as import |
| Microsoft Fabric | Lakehouses, warehouses, semantic models | Direct Lake is available for Fabric data |
| Power Platform | Dataverse, dataflows | Dataverse holds Dynamics 365 app data |
| Azure | Azure SQL Database, Azure Data Lake Storage, Azure Databricks | Cloud sources need no gateway |
| Online services | SharePoint Online lists, Salesforce, Dynamics 365 Business Central | Sign-in uses each service's own account |
| Other | Web, OData, ODBC, blank query | Generic routes for systems with no own connector |
Does every connector allow DirectQuery?
No. Import mode works with nearly every connector, because it simply copies the data into the model. DirectQuery is different. It sends a query to the source each time a reader opens a page, so the source must answer quickly and the connector must support it.
Most large databases and cloud warehouses allow DirectQuery. However, files and many online services do not. For those, import is the only choice, and scheduled refresh keeps the copy current. Microsoft's connector pages list the modes each one allows, so check there first.
Fabric adds Direct Lake for data stored in OneLake. It reads Delta tables directly, so it feels close to import while staying fresh. Also, a live connection lets a report reuse a model that already sits in the service.
When do Power BI supported data sources need a gateway?
The service runs in Microsoft's cloud, so it cannot see inside your network on its own. Therefore any source on a private network needs an on-premises data gateway for refresh. That covers office SQL Servers, file shares and files saved on a local drive.
Cloud sources usually skip the gateway. For example, Azure SQL Database, SharePoint Online and most online services refresh straight from the service. Still, some setups need a gateway even in the cloud, such as a database behind a private network rule.
Custom connectors also need a gateway for refresh in the service. In addition, check the refresh limits for your license, because they cap how often a schedule can run each day.
Checks to run before you choose a connector
A short review with the source owner avoids surprises once the report goes live.
- Is the connector listed and generally available?
- Which modes does it allow: import, DirectQuery or both?
- Will refresh need a gateway in the service?
- Which sign-in method does the source accept?
- Can the source handle the query load at peak times?
- Who owns the source and approves access?
How to confirm connector support for your own systems
Follow these steps for each system you plan to report on.
List your systems
First, write down each source, where it runs and who owns it. Include spreadsheets that people update by hand.
Check Microsoft's lists
Next, look up each one in the Power BI and Power Query connector pages, and note the modes and gateway needs.
Test a small load
Then connect in Desktop, load a sample and publish it. After that, run a refresh in the service.
Pick the route
Finally, choose the connector, mode and gateway for each source. Our consultants can run this check with you.
Power BI supported data sources: questions teams ask
How many connectors does Power BI offer?
Hundreds, and the list grows with each monthly Desktop release. Because the number changes, rely on Microsoft's connector pages rather than a fixed count. They show each connector, its release stage, the modes it allows and whether it works in Desktop, dataflows or both.
Can Power BI connect to a system with no connector?
Often, yes. Many systems offer an ODBC driver, an OData feed or a REST API, and generic connectors reach those. Otherwise, a nightly export to a database or a file in SharePoint works well. The right route depends on volume, security and how fresh the data must be.
Do Excel files refresh automatically in the service?
Yes, when the file sits in OneDrive or SharePoint Online. The service can then read it straight from the cloud. However, a file saved on a laptop or a network drive needs an on-premises data gateway, and the computer holding it must be on during refresh.
Are third-party connectors safe to use?
Vendor connectors that ship inside Desktop pass a Microsoft review first. Custom connectors that you install yourself skip that review, so your admin decides whether to allow them. In both cases, check who maintains the connector and how it handles sign-in.
Which Power BI supported data sources work best for large data?
Cloud warehouses and large databases suit big volumes, because they can answer DirectQuery fast or feed incremental refresh. Fabric lakehouses with Direct Lake are another strong option. Flat files work for smaller tables, but they slow down as the data grows.
Let's build your first dashboard together.
Tell us what you report on today. A consultant replies within 2 hours with next steps.
Trusted partners
Microsoft Solutions Partner · AWS Partner
Cloud & AI Platforms
Azure SQL, Microsoft Fabric and the data pipelines behind your reports.
AI Business Solutions
Power BI reporting on your Dynamics 365, Business Central and Power Platform data.
Security
Row-level security and Entra ID sign-in on every report and workspace.
Amazon Web Services
Power BI on your AWS data, with connectors for Redshift, Athena and S3.
View our Microsoft partner listing