Parquet Viewer
Drop a .parquet file here, or .
What this tool does
Open an Apache Parquet file and inspect it without a database, notebook, or CLI. The viewer reads the file footer to show row count, row groups, compression codec, and the full column schema with physical and logical types, then pages through the actual rows in a table. Export the page you are looking at as CSV or JSON.
When to use this tool
Quickly check what a data export actually contains — column names, types, null-heavy fields, and a few sample rows — before wiring it into a pipeline. It is also handy for verifying a file someone sent you is valid parquet, and for reading the compression and row-group layout when you are tuning file sizes.
Supported files
Uncompressed, Snappy, Gzip, Brotli, LZ4, and Zstandard column chunks are all decoded, along with dictionary, RLE, and delta encodings. Nested columns (structs, lists, maps) appear in the schema by their full dotted path and render as JSON in the table. Encrypted parquet files are not supported.
Privacy and limitations
Parsing happens entirely in your browser — the file never leaves your machine and no upload is made. Only the footer and the row groups behind the current page are decoded, so a large file opens fast, though a page whose row group is very wide still takes a moment. Exports cover the rows currently on screen, not the whole file.