Parquet Viewer | Tools Hub

or click to choose. Files stay in your browser and are never uploaded.

Try a sample:

Overview - rows, columns, size, compression ratio, and embedded pandas/Arrow/Geo metadata.
Schema - physical and logical types with repetition, including nested groups.
Row Groups - size and compression of each independently readable slice.
Columns - per-column storage, codec, encodings, null counts, and min/max from the footer statistics.
Data Preview - paginated rows, exportable to CSV or JSON.
Raw Metadata - the full parsed file footer.

Statistics (min/max/nulls) are read from the file footer, not by scanning data, so they load instantly for any file size.
Snappy and uncompressed files work immediately; gzip/zstd/brotli/lz4 load a small extra decoder on first preview.
URL loading needs the host to allow cross-origin requests. If it fails, download the file and drop it in.

Local only. Parsing happens entirely in your browser. Files and their contents are never sent to any server.

Attributions & Licenses