JSON to Parquet Converter
Convert a JSON array, a single object, or a newline-delimited JSON (NDJSON) file to Parquet in your browser, with nested objects and lists preserved exactly — never flattened.
Nested structure survives, on purpose
Unlike JSON to CSV and JSON Flattener, this tool does not flatten anything. Parquet is a typed, columnar format that can represent a nested object or a list natively, so your file's schema — including nested fields — is inferred once and written straight through. A record that had a list of objects in the original JSON has that same list of objects in the resulting Parquet file, not a flattened or JSON-text approximation of it.
Schema inference
Column types (including nested STRUCT and LIST types) are inferred directly from your file by the same local database engine every tool on this site uses, and shown in the preview before you download — never a blind conversion.
Browser-local processing
Conversion runs entirely inside DuckDB, running as WebAssembly in your browser. Your file is not uploaded. Parquet is a binary format that a spreadsheet program does not open directly, so the formula-injection safeguard this site's CSV exports apply does not apply here — it is a CSV-specific, spreadsheet-specific concern, not a general one.
Large files
This tool warns you above roughly 100 MB, warns more strongly above roughly 500 MB, and declines to open a file above roughly 2 GB rather than risk crashing the tab.
Related tools
- JSON to CSVConvert JSON arrays to CSV, flattening nested fields.
- Parquet ViewerInspect the schema, rows and real file structure of a Parquet file without uploading it.
- CSV to ParquetComing soonConvert CSV to compressed, typed, columnar Parquet.
- JSON FlattenerFlatten nested JSON into flat key paths, previewed before export.