Data Converter

About the Data Converter

The Data Converter turns a dataset in one format into another — CSV in, Parquet out; Excel in, JSONL out; whatever pairing you need. Everything runs in your browser using PapaParse, SheetJS, hyparquet, and Apache Arrow JS, so the file never leaves your device. Drop in a file up to 50 MB, eyeball the inferred column types in the preview, then export to the format your next tool actually wants.

What you can convert between

• Input formats: CSV, TSV, TXT, JSON, JSONL/NDJSON, Excel (.xlsx, .xls), Parquet, Feather/Arrow IPC.
• Output formats: CSV, TSV, JSON (array of objects), JSONL, Excel, Parquet, Feather/Arrow.
• Auto type inference for each column so numeric, boolean, date, and string values survive the round trip where the target format supports them.
• Preview the first 50 rows before downloading, so you catch parsing surprises (wrong delimiter, header row, encoding) early.
• Custom output filename, or leave it blank for an auto-named export based on the input.

Common uses

• Shrinking a giant CSV into a Parquet file before loading it into a notebook.
• Flattening an Excel workbook into JSON for a web API or static site.
• Converting JSONL exports from an API into a tidy CSV for a spreadsheet.
• Handing collaborators a Feather/Arrow file that loads instantly in pandas or R.
• Cleaning up oddball delimiters by re-exporting through CSV/TSV.

Tips

• Parquet and Feather preserve column types best — use them when downstream code cares about dtypes.
• If your CSV has unusual encoding or quoting, run a quick preview before downloading to confirm columns split correctly.
• For very wide datasets, Parquet typically yields the smallest file; for human-readable diffs, stick with CSV or JSONL.
• R’s .rda and Python’s .pkl aren’t supported in-browser — export to CSV or Parquet first.

Everything happens locally. No upload, no cloud step, no waiting for a server — and no chance of a private dataset ending up somewhere it shouldn’t.

Common questions

How do I convert CSV to JSON?

Drop the CSV onto the converter, pick JSON (or JSONL for line-delimited), and download. The input format is auto-detected, headers become keys, and types are handled sensibly.

Can I convert Excel files to CSV or JSON without Excel?

Yes. Drop the .xlsx file in and convert straight to CSV, JSON, or any other supported format, no spreadsheet software required.

What is Parquet and why would I convert to it?

Parquet is a compressed columnar format that data tools like pandas, Spark, and DuckDB love: smaller files, faster analytics. This is one of the few browser tools that reads and writes it, along with Feather/Arrow.

Is it safe to convert files with real customer data?

Yes, because nothing uploads. The conversion runs entirely in your browser, so datasets full of names, emails, and transactions never leave your machine, which is the only acceptable way to convert them.

Is there a file size limit?

No artificial one. Your device does the work, so files far larger than most online converters accept process fine, and there is no upload wait before conversion starts.

Related tools