1. 01Choose the source and target formats; the two cannot be the same.
2. 02Paste the data, or receive a .json, .yaml, .yml, .xml or .csv file handed over inside the site.
3. 03About 200 ms after you type, it parses and converts automatically; on failure the parser error is shown.
4. 04Copy the result, or download it with the matching extension.

• JSON uses the native JSON.parse, so comments, trailing commas and unquoted keys — anything that is not standard JSON — fail.
• YAML uses js-yaml; XML is validated first with fast-xml-parser, whose errors carry a line and column, and attributes are then kept with an @_ prefix while text values are trimmed.
• CSV uses the first row as the header, skips blank lines and enables dynamic typing; the whole text is trim()ed before parsing, so leading and trailing blank lines are not preserved.
• The CSV output needs to be mappable to an array of objects; a nested object or array is JSON.stringifyed inside the cell rather than spread across columns.
• If the XML output value is not a single root object it is wrapped in <root> automatically, and array elements use <item>, so a round trip across formats cannot guarantee that the original structure names survive.

Parsing and format conversion both happen locally in the browser; a file is read as text only and is never sent to a server.