QVD to Parquet Converter
Turn a QlikView or Qlik Sense .qvd file into an Apache Parquet file, ready for pandas, DuckDB or Spark. The conversion runs in your browser: your file is never uploaded.
How your data is handledUpload QVD File
Drag and drop or click to select
Why convert a QVD to Parquet
A QVD only opens in Qlik, which is a problem as soon as the same data is needed in a notebook, a warehouse or a data lake. Parquet is the format that side of the world speaks: columnar, typed and compressed, read natively by pandas, Polars, DuckDB, Spark, BigQuery, Snowflake and Athena.
Compared with exporting the QVD to CSV, Parquet keeps the column types, so no one has to guess whether a field was a number or text, and the file is typically several times smaller: a table that becomes a 178 MB CSV lands at about 60 MB as Parquet. Queries that touch a few columns only read those columns.
Qlik can write Parquet with a STORE statement, but only from a licensed environment with the QVD reloaded. This converter does it from a browser tab, reading the QVD format directly.
How to convert a QVD file to Parquet
- 1
Open the QVD file
Drop your .qvd file onto the converter or pick it from your computer. Your browser reads it directly, so nothing is uploaded and there is no upload size limit.
- 2
Review the schema
The dashboard lists each field with its number of distinct values and whether it holds text or numbers, which is what determines the Parquet column type.
- 3
Click Export Parquet
An Apache Parquet file is written in your browser, in row groups of 100,000 rows, and saved to your downloads folder ready for pandas, DuckDB, Spark or a data lake.
What you get in the Parquet file
- Typed columns: INT32 for whole numbers, DOUBLE for decimals, UTF-8 strings for text and mixed fields.
- Row groups of 100,000 rows, so readers can skip past the parts of the file a query does not need.
- Real nulls, rather than empty strings, for values missing in the QVD.
- Original field names, kept as the Parquet schema's column names.
QVD to Parquet: frequently asked questions
How do I convert a QVD file to Parquet?
Open the .qvd file in this converter and click Export Parquet. The Apache Parquet file is written in your browser, so you need neither Qlik nor a Python environment.
Why convert QVD to Parquet instead of CSV?
Parquet keeps column types, so numbers stay numbers and nothing has to be re-parsed on load. It is columnar and compressed, so the file is usually several times smaller than the same data as CSV and much faster to query in pandas, DuckDB, Spark or a data lake.
Which Parquet types does the converter use?
Each column is inspected and given the narrowest type that fits every value: INT32 for whole numbers within the 32-bit range, DOUBLE for other numbers, and a UTF-8 string column for text or mixed content. Missing values are written as Parquet nulls.
Can I read the exported file in pandas or DuckDB?
Yes. It is a standard Apache Parquet file: pandas.read_parquet('file.parquet'), DuckDB's read_parquet, PyArrow, Polars, Spark and cloud warehouses all read it directly.
How large a QVD can be converted to Parquet?
Millions of rows are fine: a 1.2 million row table exports in about 20 seconds, and files with tens of millions of rows work as long as your browser has the memory. Parquet export is slower than CSV because the data is typed and encoded per column.
Is the QVD to Parquet converter free?
Yes, free and without sign-up. The file is converted on your device and never uploaded, so nothing is stored on our side.