How to Open a Parquet File Without Spark or Python

The problem

Someone hands you a .parquet file and you just want to see what's inside — the schema, a few rows, maybe a row count. Installing Spark for that is absurd, and even pip install pandas pyarrow is a detour when you're on a locked-down laptop or someone else's machine.

Here are the realistic options, fastest first.

1. A browser-based viewer (zero install)

Drag the file into the Parquet Viewer on this site. It reads the metadata footer and only the rows you look at, so gigabyte files open in under a second — and because it runs on WebAssembly in your browser, the file is never uploaded anywhere.

2. DuckDB CLI (one binary)

If you live in a terminal, DuckDB is a single ~40 MB binary:

SELECT * FROM 'file.parquet' LIMIT 10;
DESCRIBE SELECT * FROM 'file.parquet';

3. Python with pandas or Polars

The classic route when you already have a Python environment:

import pandas as pd
df = pd.read_parquet("file.parquet")   # needs pyarrow installed

4. VS Code extensions

Several extensions render Parquet as a table inside the editor. Convenient if the file is already in your workspace, though large files can be slow.

5. Converting to CSV

When the real goal is "get this into Excel / Sheets", skip the viewer and convert Parquet to CSV directly — then open the CSV anywhere.

Which should you use?

Situation	Best option
Quick peek, no installs allowed	Browser viewer
Repeated ad-hoc SQL on many files	DuckDB CLI or SQL Workbench
Already inside a Python workflow	pandas / Polars
Handing data to a spreadsheet user	Convert to CSV