JSON to CSV Converter

CSV Viewer & Validator renders delimited text as a table and points out where it is inconsistent. It detects the delimiter from the first line — comma, semicolon or tab — and shows the row count, the column count and any row whose number of columns differs from the header. The parser is a real state machine, so quoted fields, escaped double quotes and line breaks inside a quoted cell are handled correctly; a stray quote in the middle of a field is kept as-is rather than raising an error. It only displays the data — it does not fix, sort or reformat it. To turn the rows into objects for code, use CSV to JSON; to strip repeated rows, use CSV Duplicate Row Remover. Input is limited to 2 MB, 50,000 rows and 250,000 cells, and the file is read entirely in the browser.

CSV to JSON Converter turns the rows of a delimited file into JSON, using the same parser as CSV Viewer (delimiter guessed from the first line — comma, semicolon or tab). If you mark the first row as a header, the output is an array of objects keyed by that row; without a header it is an array of arrays. Every value comes out as a string — 42, true and blank cells are all quoted text, with no type guessing. An empty header becomes column_1, column_2 and so on; a repeated header gets _2, _3 appended; the risky keys __proto__, constructor and prototype are renamed. A row shorter than the header fills the missing keys with empty strings; a row with extra cells has them dropped. Choose two-space pretty output or a compact one. Input is capped at 2 MB, 50,000 rows and 250,000 cells, and the file is read in the browser.

CSV Duplicate Row Remover deletes rows that repeat and keeps the first of each, leaving every other row in its original position — it never sorts. Two rows are duplicates when their parsed cells match across the whole row. Case-sensitive comparison is on by default; untick it to treat "Anna" and "anna" as the same. It compares parsed values, so a quoted "a" and a bare a match, but a cell with an extra leading or trailing space does not match one without — quoting and stray whitespace can keep near-identical rows apart. The first row is treated as a header by default: kept untouched and excluded from the comparison, unless you tick the box to include it as data. The delimiter is auto-detected from the first line or chosen manually, the output uses LF line endings, and a count of removed rows is shown. Unlike CSV Viewer it rewrites the file, and unlike CSV to JSON the result is still CSV. Limits: 2 MB, 50,000 rows, 250,000 cells.

Reorder the rows of a CSV by one column you choose, entirely in your browser. Paste your CSV, pick the delimiter (comma, semicolon, tab, or auto-detect), say whether the first row is a header, then read the columns and sort. You choose the sort type explicitly — text, number, or ISO date — because guessing is unreliable for IDs, zero-padded values and locale date formats. Text sorting is case-sensitive and deterministic; number sorting accepts standard numeric literals and rejects thousands separators, currency and integers beyond the JavaScript safe range; date sorting accepts only YYYY-MM-DD and ISO 8601 timestamps with a time zone. Empty cells and values that do not match the chosen type are always placed last, ascending or descending, and rows with an equal sort key keep their original order (stable sort). The header row stays in place. Quoted cells, embedded commas and embedded newlines are preserved through the shared RFC 4180 parser and serializer; malformed CSV is rejected rather than repaired. Nothing is uploaded or stored. This version sorts by a single column; it is not a spreadsheet, it does not remove or select columns, and it does not remove duplicate rows.