Parquet viewer

A free Parquet viewer that just opens the file

Apache Parquet is a columnar binary format built for fast reads at scale — which also makes it unreadable as plain text. Double-clicking a .parquet file gets you a wall of binary in a text editor; actually seeing the rows normally means opening Python, installing Pandas or PyArrow, and writing pd.read_parquet(...) just to check what's in a file someone emailed you or a data pipeline dropped in a folder.

Sarvak skips that. It's a free, native desktop app — open a .parquet file and it renders the schema and rows into a table immediately, the same way it opens Markdown, JSON, or a dozen other formats.

Download Sarvak Windows · Linux · macOS — free

What the Parquet viewer does

  • No Python required. Reads the file directly — no Pandas, PyArrow, or DuckDB install, no virtual environment to set up first.
  • Handles compressed files. Snappy, GZIP, and ZSTD Parquet pages are decompressed automatically, covering the defaults from Spark, Pandas, and DuckDB — not just the uncompressed case.
  • Table view of schema and rows. Columns render as a table, with the first 1,000 rows shown for a quick look at real data rather than a hex dump.
  • Instant and offline. Nothing is uploaded — the file is read from local disk, rendered client-side, and never leaves your machine.

How to open a .parquet file

  1. Download Sarvak for Windows, macOS, or Linux (under 6MB, no install wizard beyond the standard installer).
  2. Drag the .parquet file into the window, or use File > Open.
  3. Browse the table. Use Ctrl/Cmd+F to search within what's rendered.

Also opens the rest of your data files

If you're regularly digging through data exports, Sarvak's format list also covers SQLite databases with a read-only SQL query box, CSV/TSV with a searchable table view, and JSON/YAML/XML with folding — one app instead of a different tool per file type.

FAQ

How do I open a .parquet file without Python or Pandas?

Download Sarvak, then open the .parquet file directly from disk (drag it in, or use File > Open). Sarvak reads the file's schema and rows into a table view immediately — there's no Python interpreter, Pandas/PyArrow install, or Jupyter kernel involved.

Does Sarvak support compressed Parquet files (Snappy, GZIP, ZSTD)?

Yes. Sarvak decompresses Snappy, GZIP, and ZSTD Parquet pages, which covers the defaults used by Spark, Pandas, and DuckDB, not just uncompressed files.

Is there a size limit on Parquet files Sarvak can open?

Sarvak's Parquet viewer reads the whole file into memory client-side and shows the first 1,000 rows, so it's built for the kind of Parquet exports and samples you'd inspect by eye, not multi-gigabyte data-lake partitions.

Does opening a Parquet file in Sarvak upload it anywhere?

No. Sarvak is a native desktop app that reads the file from local disk. Nothing is uploaded to a server or cloud service to render the table.