.parquet — Apache Parquet
Data application/vnd.apache.parquet
What a .parquet file is
A columnar storage format built for analytics. Because it stores each column together, queries that touch a few columns of a huge table read far less data.
How to open it
Pandas, DuckDB, Spark, or the DuckDB CLI for a quick look.
Worth knowing
Typically 5–10× smaller than the equivalent CSV, and much faster to query. Not human-readable.
How to identify a .parquet file
A file's extension is only its name — renaming a.png to a.parquet changes nothing inside. What software actually reads is the magic number, the first few bytes of the file.
| Signature | 50 41 52 31 |
| MIME type | application/vnd.apache.parquet |
| Category | Data |
ASCII "PAR1" at both the start and the end of the file.
# what is this file really?
file mystery.parquet
# look at the first bytes yourself
xxd mystery.parquet | head -2
head -c 16 mystery.parquet | xxd
# Windows PowerShell
Get-Content mystery.parquet -AsByteStream -TotalCount 16 | Format-HexFrequently asked questions
What is a .parquet file?
A columnar storage format built for analytics. Because it stores each column together, queries that touch a few columns of a huge table read far less data.
How do I open a .parquet file?
Pandas, DuckDB, Spark, or the DuckDB CLI for a quick look.
Is .parquet safe to open?
Yes. A .parquet file is data, not code — opening one cannot execute anything by itself.
Can I change a file's extension to convert it?
No. Renaming changes the label, not the bytes. The one place it appears to work is between formats that genuinely share a container — renaming .docx to .zip works because a DOCX is a ZIP. Everything else needs a real converter.
Other data formats
- .csv — Comma-Separated Values
- .json — JavaScript Object Notation
- .yaml — YAML Ain't Markup Language
- .sql — SQL Script