Convert a small CSV table for an import, prototype or data exchange. Values remain strings, so identifiers such as 00127 keep their leading zeros. Processing is local, and quoted commas or embedded line breaks are handled as part of their fields.
Preserve the value before assigning a type
A CSV field that looks like a number may be a postal code, product identifier or telephone extension. This converter keeps values as strings rather than deciding that numeric-looking text should become a number. The value 00127 therefore remains 00127 inside a JSON string. Dates and true or false text are also not automatically interpreted.
Your receiving application may require numbers or booleans for specific fields. Make those changes deliberately after conversion using that application's schema. Avoid converting every numeric-looking value indiscriminately: the same column may contain leading zeros or values that are identifiers rather than quantities.
Example: preserve a product code and a comma
Consider a table with product code 00127 and a description reading Blue mug, large. The description contains a comma, so the CSV field must be enclosed in double quotes. After conversion, check that the code still has five characters and the description is one value rather than two columns.
A field may also contain a line break when the whole field is quoted. That means a visible new line in the pasted text is not always the start of another record. The parser handles quoted fields; splitting the input at every comma or line break would not be a safe substitute.
Fix malformed CSV at the source
If conversion reports inconsistent columns, inspect the record near the problem. A comma in an unquoted description can create an extra field. A missing closing quote can make several later lines appear to belong to the same field. Within a quoted field, a literal double quote is represented by two consecutive double quotes.
Choose the separator used by your source: comma, tab or semicolon. If everything appears in one column, check that setting before editing the data. Keep the first-row header option enabled when that row names your columns; disable it when the first row is ordinary data.
Check a conversion before an import
Review the first and last records, an empty field, a value containing punctuation and an identifier with leading zeros. Confirm that the receiving application expects the resulting JSON structure and field names. Valid JSON can still be the wrong shape for a particular application.
The limits are 1 MB of input, 10,000 rows and 200 columns. For a larger export, make smaller complete tables rather than cutting text in the middle of a quoted field. Keep the original export so a correction does not require reconstructing lost data.
Frequently asked questions
Does this upload my CSV?
The text is processed in your browser and is not uploaded for conversion. Copy or download the result before reloading the page. The tool does not provide persistent storage for your table.
Why are numbers in quotation marks?
Values remain strings intentionally. CSV does not carry a reliable type schema, and automatically converting numbers can remove meaningful leading zeros. Apply required types later with knowledge of the destination's fields.
Can I convert JSON back to CSV?
Use the JSON to CSV tool for that direction. It supports flat records and rectangular arrays, with formula protection enabled by default for CSV exports intended for spreadsheet use.