The .pbix file.
Also .pbit, which are read the same way.
- Really a
- ZIP container
- Holds
- A compressed columnar model plus DAX measures
- Import mode
- Contains the data
- DirectQuery mode
- Contains no data, only the schema and measures
- Template .pbit
- Model and measures with the data deliberately removed
- Read here as
- Extracted. The model's tables. Measures come from a .pbit, not from a .pbix.
- Status
- Ready
- Fidelity
- Data extracted
- Largest file
- 25 MB free, 250 MB once a project is bought
What it actually is
The tables inside the report are read out of its model, along with the measures that were defined on top of them, so your own definitions are used rather than re-invented.
A .pbix is a ZIP container. Inside it sit the report layout, and a compressed columnar data model holding the tables that were imported, plus the DAX measures defined on top of them.
Reading the measures matters as much as reading the rows. A measure named Margin is somebody's decision about what margin means in this business, written down and used in production. Re-inventing that definition when the file already states it is how two reports come to disagree.
The exception is a report in DirectQuery mode, which holds no data of its own and queries a database at render time. That file contains a schema and measures and nothing else, which is a fact worth being told rather than discovering through an empty result. A .pbit template is the same situation by design.
Worth knowing: A report that queries a live database holds no data of its own. We read the schema and the measures and tell you which query to export.
Upload one
The analysis runs before there is anything to pay for.
You see what it found, and the evidence behind it, first.
No account needed to start. You only pay when you like what you see.
What goes wrong, and what is done about it
2 named failures, each with the repair.
A file with no rows in it
A DirectQuery report is a set of queries pointed at a warehouse. It looks like any other .pbix, is often large, and contains none of the numbers on screen when it is open.
What is done: Nothing detects the storage mode. A DirectQuery report yields no tables, which arrives as an empty result rather than as an explanation, so check whether the report imports its data before uploading it.
Measures that do not mean what their name says
A DAX measure can filter, override context and reference other measures. Two measures called Revenue in the same model routinely compute different things, and the report shows only the name.
What is done: Not from a .pbix, which yields its tables and nothing else. Save the report as a .pbit template and upload that beside it: the measure expressions are read out of a template, so a definition can be checked rather than assumed from a label.
Questions people ask
About this format, not about the product.
Do I need Power BI installed?
No. The .pbix container is read directly and the tables inside its model come out, without the desktop application. The measure definitions do not come out of a .pbix, though: save the report as a .pbit template and upload that too if you want what a measure actually computes.
Why would I analyse a report I already have?
Usually because you want a question answered that the report was not built to answer, and the data is trapped behind its visuals. The model inside is the dataset, and everything on this platform works on it.
What about Tableau?
The same job, a different container. A packaged .twbx carries its extract and reads as data; a bare .twb points at a source and yields calculated field definitions only.
Keep reading
The jobs people do with this file, and the formats beside it.