Parquet File Viewer
Free Parquet file viewer. No upload, parsed locally.
Loading tool...
What is Parquet File Viewer?
The Parquet File Viewer opens Apache Parquet files directly in your browser and answers the first questions a data engineer asks: how many rows, how many row groups, how large is the file, and what columns does it hold. The schema table lists every column with its physical and logical type, the preview grid decodes the first hundred rows with nested values shown as JSON snippets, and one click exports the preview to CSV or JSON. Parsing runs locally through hyparquet, so the file never leaves your device.
Key Benefits
- Instant metadata: rows, row groups and file size
- Full schema with physical and logical column types
- Scrollable hundred-row preview with readable nested values
- CSV and JSON export of the previewed rows
- Completely local parsing — safe for confidential data files
Common Use Cases
- •Inspecting a data drop from a vendor before loading it
- •Checking column names and types before writing ETL
- •Sampling a large Parquet dataset without spark or duckdb
- •Pulling a CSV slice for a quick chart or bug report
- •Teaching what Parquet row groups and schemas look like
How to Inspect a Parquet File
- Open the file: Drop a .parquet file onto the box or pick one with the file button. The whole file is read locally — nothing is uploaded.
- Read the summary: The summary cards show total rows, row-group count and file size, which tell you how heavy the file is before any tool loads it.
- Check the schema and preview: The schema table lists every column with its physical and logical type, and the preview grid shows the first hundred rows with nested values rendered as JSON snippets.
- Export what you need: Export the previewed rows as CSV for spreadsheets or JSON for notebooks and bug reports, then close the file to drop it from memory.
Key Features
- Metadata summary: rows, row groups and file size
- Full column schema with physical and logical types
- Scrollable preview of the first hundred rows
- One-click CSV or JSON export of the preview
- Fully local parsing with drag-and-drop support
Frequently Asked Questions
Is my Parquet file uploaded anywhere?
No. The file is read with the browser File API and parsed locally by hyparquet, a pure-JavaScript Parquet reader. No bytes leave your device, there is no server component, and closing the file releases it from memory.
How much of the file does the preview show?
The preview decodes the first hundred rows, which is enough to check column names, sample values and data quality. The summary cards still report the true total row count and row-group count for the whole file, so you know exactly how big it is.
Which Parquet compressions are supported?
Uncompressed and Snappy-encoded columns are handled out of the box, which covers the vast majority of files written by Spark, pandas, duckdb and Arrow. Files using exotic codecs or deeply nested structures may fail with a message naming the problem.
How do I get the whole file into CSV?
This viewer exports the previewed rows to keep the browser responsive on large files. For a full conversion, export the preview to verify the schema, then convert the complete file with a local tool such as duckdb or pandas — the column types you confirmed here tell you exactly what to expect.