How to Open a Parquet File in Your Browser

1. Choose your local Parquet file

Open the Parquet viewer and choose a .parquet file from your device, or drop it onto the file entry area. The detector checks the header and footer before handing the file to the reader. You do not need an account or a database connection. Files are processed locally with DuckDB-Wasm; loading the website and its engine assets still requires a network connection.

2. Inspect schema and browse records

Check the schema and row count to understand what the dataset contains. Browse table pages, search values, filter records and sort columns to compare results. Data is materialized in a temporary local DuckDB table. There is no SQL editor in this release, so use the provided table controls for inspection.

3. Export rows or keep the original

CSV export supports the current page or all filtered and sorted matches, with at most 100 rows per page and a 50 MiB generated-export cap. Select All matching rows to export beyond the current page. Download original keeps the original bytes unchanged. Closing or reloading clears the local session; downloaded files remain on your device.

If your Parquet file does not open

The input cap is 200 MiB, but compressed data can exhaust browser memory earlier. Encryption and some codec variants are unsupported; nested and binary variants need broader compatibility checks. Try a smaller extract from your source tool. No upload or remote processing fallback is used.

Questions and answers