CSV to JSON Converter

A CSV to JSON converter transforms rows and columns from a comma-separated values file into structured JSON. With a header row, each column name becomes an object key and each following row becomes a JSON record that applications, APIs, and databases can consume directly.

Yes. ToolsSonic parses pasted CSV and local files entirely in your browser using JavaScript. The converter does not upload the input, store it in localStorage, or send rows to a conversion API. This makes it safe for internal data, client exports, and confidential spreadsheets.

Yes. The parser recognizes RFC 4180 quoted fields, escaped double quotes written as two consecutive double quotes, embedded commas, and line breaks inside a quoted cell. This handles CSV exports from Excel, Google Sheets, and most database tools without errors.

You can auto-detect or manually select comma, semicolon, tab, pipe, and a custom one-character delimiter. European CSV files commonly use semicolons because commas are used as decimal separators — just select semicolon or auto-detect and the converter handles it correctly.

Yes. Turn off the first-row-contains-headers option and the converter creates generated keys such as column_1, column_2, and column_3. If your first row contains field names, keep the header option enabled for readable, meaningful JSON keys.

Yes. Type inference converts numeric strings to JSON numbers and true or false strings to JSON booleans. You can also convert empty cells to null. Disable type inference when every value must remain a string, such as product IDs or order numbers that contain leading zeroes.

Four formats are available: array of objects (the most common format for REST APIs and JavaScript), objects keyed by a selected column (useful for lookup tables), column arrays (one array per field, useful for charting libraries), and JSON Lines or NDJSON (one JSON object per line, useful for streaming, logging, and MongoDB imports).

Yes. Enable the nested dot keys option to turn column headers such as user.name and user.email into a nested user object containing name and email properties. This is useful when importing CSV into a document database or an API that expects nested payloads.

Export your spreadsheet as CSV from Excel (File → Save As → CSV) or Google Sheets (File → Download → CSV). Then drag the file into the ToolsSonic converter or click Upload file. The converter reads the file locally, shows a live table preview, and generates JSON you can copy or download.

Keep the duplicate-header normalization option enabled to receive unique keys such as email and email_2. If duplicates are accidental, fix the source spreadsheet before using the JSON in production so downstream code has stable and predictable field names.

An array-of-objects output is one valid JSON document with all records inside square brackets — suitable for REST APIs, fetch requests, and JavaScript arrays. JSON Lines writes one independent JSON object per line with no surrounding array — suitable for streaming processors, log pipelines, MongoDB mongoimport, and command-line tools like jq.

Yes. Either select Tab from the delimiter dropdown or use Auto-detect, which recognises tab-separated values automatically. TSV files exported from databases, Excel, or BI tools convert to JSON the same way as standard CSV files.

The parser follows RFC 4180, so a quoted field like "Smith, John" stays one value, escaped quotes are honoured, and line breaks inside quotes do not split rows — the cases that break naive converters.

Four: an array of objects (the default for APIs), an object keyed by a chosen column, per-column arrays for charting libraries, and JSON Lines (NDJSON) for tools like mongoimport and jq.

Yes. Type inference turns 42 into a number and true into a boolean, and can map empty cells to null. Disable it when codes with leading zeroes or phone numbers must stay strings.

Yes. With dot-key nesting enabled, headers like user.name and address.city produce nested objects instead of flat keys.

Yes — save the sheet as CSV from Excel or Google Sheets first, then convert. Quoted commas and line breaks from spreadsheet cells are handled correctly.

Everything runs locally in your browser with no upload, no account, and no logging. You can confirm in your browser DevTools Network tab that no request carries your data.