How to open a Parquet file on a Mac

By Mykola-Bohdan Vynnytskyi · Updated November 7, 2026

macOS has no built-in app for Parquet, so a double-click in Finder will not show you the data. Install a viewer that registers the .parquet file type, or read the file with pandas or the DuckDB command line. All of these run on your Mac, and the file is not uploaded.

Parquet is a binary, column-oriented format. Rows are laid out by column and usually compressed, and the schema sits in a footer at the end of the file. TextEdit and Quick Look cannot help, because something has to decode the file first.

Open it in Parquet Reader

1. Install Parquet Reader from the Mac App Store. Single files are free and no account is needed.
2. Double-click the .parquet or .pq file in Finder. If another app already claims the file type, right-click the file, choose Open With and pick Parquet Reader. Inside the app, Open File or ⌘O does the same.
3. The data appears as a table. The card at the top shows the row and column counts, and the pager moves through the rows.
4. Right-click a column header to sort, filter or choose View Statistics.
5. When a look is not enough, use the SQL Query panel. The file is available as parquet_table, and the first query is already filled in: SELECT * FROM parquet_table LIMIT 100.

Other ways to open it

You do not need an app if Python or a terminal is already part of your day.

pandas

pip install pandas pyarrow

import pandas as pd

df = pd.read_parquet("orders.parquet")
print(df.shape)
print(df.head())

With no other arguments, pandas loads every row and column into memory. On a large file, pass columns=[...] to read only what you need.

DuckDB command line

brew install duckdb
duckdb -c "SELECT * FROM 'orders.parquet' LIMIT 10"

┌──────────┬──────────┬─────────┬────────┬────────────┐
│ order_id │  status  │ country │ amount │ order_date │
│  int64   │ varchar  │ varchar │ double │    date    │
├──────────┼──────────┼─────────┼────────┼────────────┤
│     1001 │ paid     │ UA      │   49.9 │ 2026-01-03 │
│     1002 │ paid     │ DE      │  120.0 │ 2026-01-04 │
│     1003 │ refunded │ UA      │   15.5 │ 2026-01-09 │
│     1004 │ paid     │ PL      │   NULL │ 2026-02-01 │
│     1005 │ pending  │ DE      │  80.25 │ 2026-02-02 │
└──────────┴──────────┴─────────┴────────┴────────────┘

In a browser

Browser viewers built on DuckDB-WASM decode the file inside the tab and upload nothing. Whether your company allows a web tool for work files is a separate question.

What the free version does not do

Free opens files of any size, and you can keep several open as tables and join them in SQL. Opening a whole folder of Parquet files as one dataset is a Pro feature, covered in the guide to partitioned folders.