Nothing is uploaded. When you paste text or open a file the converter reads it locally with the browser’s File API, parses it into memory, flattens and serialises it to CSV on your device in JavaScript, and builds any .xlsx export locally too. There is no server, API, database or tracking in the processing path, so an API dump, customer export or any other sensitive data stays on your machine. The tool works offline once cached, and closing the tab discards everything. The one case where data touches the network is if you paste a remote URL for the tool to fetch — and even then the response is downloaded straight into your browser and never sent anywhere else. The only optional network use anywhere on the site is consent-gated advertising, which never sees your file.

CSV is flat, so the converter has to project JSON’s nesting onto columns. With flattening on (the default), a nested object is walked recursively and each leaf becomes its own column named by the dot path — {“address”:{“city”:“NYC”,“zip”:“10001”}} produces address.city and address.zip. Turn flattening off and each nested object is instead written whole as a JSON string in one cell. Arrays are always serialised as a JSON string (for example tags becomes [“admin”,“user”]) rather than being exploded into extra rows, which keeps one input record on one output row. The header is the union of every key across all records in first-seen order, so records with different shapes still align, and any record missing a column simply gets an empty cell there.

Besides copying or downloading CSV, the tool can build a genuine .xlsx workbook in the browser (via a bundled spreadsheet engine) with a bold header row and auto-sized columns. For plain CSV, comma is the default; choose tab for a .tsv that pastes cleanly into spreadsheets, semicolon for European locales where the comma is the decimal separator, or pipe for pipe-delimited pipelines. Two options make Excel behave: switch line endings to CRLF, and enable the UTF-8 BOM so Excel opens accented and non-Latin characters correctly instead of as mojibake. You can also force every field to be quoted, choose how null and missing values appear, and write booleans as 1 / 0 for numeric imports.

The parser accepts the standard array of objects (each object a row), a single object (one row), an array of arrays where the first inner array is treated as the header, an array of primitive values (a single value column), and NDJSON / JSONL where each line is its own JSON object — a mode it can auto-detect from an uploaded file. Once the columns are detected you get a column picker: click a header to include or exclude it from the output, and use the arrows to reorder columns left or right, with a one-click reset. The live preview and the row / column / size stats update as you change options, and only the selected columns are written to the CSV, TSV or .xlsx you export.