Parquet Lens

A fast viewer for Parquet, Arrow IPC and Feather v2 datasets of any size.

Open a .parquet, .arrow, .arrows, .ipc or .feather file and read it as a table: the schema and row count appear instantly from file metadata, rows decode on demand, and search streams the whole file.

Features

• Metadata-first open: schema, row count and row groups appear before any data decodes, whatever the file size.
• Virtualized table with exact type rendering: int64 and decimals at full precision, timestamps with their unit and timezone, nested lists, structs and maps as JSON.
• Whole-file search with match navigation and case toggle.
• Schema view: logical and physical types, nullability, codecs, encodings, and the footer's per-column null counts and min/max.
• Row detail pane with pretty-printed values and one-click copy; copy visible rows as CSV.
• All parquet codecs (snappy, gzip, zstd, lz4_raw) plus lz4- and zstd-compressed feather batches. Arrow IPC file and stream, including dictionary-encoded columns and legacy v4 framing.
• Works as an "Open With..." editor and as the default for these binary formats.

Privacy

Runs entirely locally. No network calls. No credentials. Ever.

The Lens family

One grammar, every IDE. JSONL Lens brings the same table, search and detail experience to JSON Lines files in VS Code. On JetBrains IDEs, the same engines ship as Parquet Lens, JSONL Lens and Log Lens.