CSV ↔ JSON Converter
Convert between CSV and JSON in your browser with RFC 4180 quoting, custom delimiters, NDJSON, dot-path flattening, and safe type inference. Zero uploads.
CSV options
Off keeps leading zeros and long IDs as strings.
Drop a .csv, .tsv, .json, or .jsonl file
About this CSV ↔ JSON converter
Convert between CSV and JSON in either direction. Supports RFC 4180 quoting, custom delimiters, header rows, type inference, NDJSON, and dot-path flattening for nested objects. Everything runs locally in your browser.
CSV → JSON
Paste a CSV (or drop a .csv / .tsv file) to get a JSON array. Pick a delimiter, toggle whether the first row is a header, and decide if you want numbers and booleans inferred. Type inference is off by default to protect long IDs and leading zeros.
JSON → CSV
Paste a JSON array of objects, an array of arrays, or NDJSON. Nested objects are flattened with dot paths by default; switch to JSON-string columns if you'd rather keep them intact. Choose LF or CRLF and whether to quote every field.
RFC 4180
Our parser follows RFC 4180: fields with commas, quotes, or newlines are wrapped in double quotes; embedded quotes are escaped by doubling them. Empty trailing fields are preserved.
FAQ
- Is type inference safe for IDs?
- Off by default. Turning it on will coerce things that look like numbers, including long IDs and values with leading zeros — keep it off if that matters.
- How are nested objects represented in CSV?
- Two strategies: flatten with dot paths (`address.city`) or keep them as JSON strings in a single column. Pick the one that matches what your downstream tool expects.
- Is NDJSON supported?
- Yes. Set the JSON source to NDJSON / JSONL when each line is its own JSON value, or pick NDJSON as the JSON output mode.
- Are my files uploaded?
- No. Parsing and stringifying happen entirely in your browser via FileReader and Blob APIs.