Use CSV and TSV without hiding data-shape changes
This tool helps you replace CSV separators with tabs without breaking quoted values, or write TSV rows as correctly quoted CSV. It is intended for spreadsheet imports, database tools, terminal output, and data exchange. Paste a small representative sample first, choose the direction, and compare the visible result with the source before you use a larger file. The page never sends the working data to NexaCurrent, so ordinary conversion stays inside the current browser tab.
Try rows containing a comma in a company name, a quoted phrase, an empty cell, and a multiline note. The result area reports the number of data rows and lists the headings it read. Review quoted values, empty cells, embedded line breaks, and any field that contains a comma or tab before using the result. If the two formats describe data differently, the page explains the exact compromise instead of simply announcing that the conversion succeeded. That review step matters whenever the result will feed another application.
Choose the separator direction and download encoding
Choose CSV to TSV when commas inside ordinary text make a tab-separated table easier to paste into another program. Choose TSV to CSV when the destination expects comma-separated data. The converter parses the source into rows and fields before changing separators, so a comma, tab, quotation mark, or line break inside a quoted value remains part of that value.
Choose ordinary UTF-8 for most modern applications. Use UTF-8 with BOM only when a spreadsheet program otherwise opens accented or non-Latin text incorrectly. The marker is added to the downloaded file, not the visible result. The converter does not change dates, formulas, numbers, or empty fields into guessed data types.
Verify headings, rows, and quoted values
Compare the listed headings and row count with the source before downloading. Both formats contain text fields only; the converter does not infer dates, numbers, formulas, or null values. A successful separator change means that field boundaries were parsed and rewritten consistently; it does not mean that a spreadsheet formula, date, identifier with leading zeros, or private import rule has been interpreted.
Pay special attention to the first and last rows, empty trailing cells, multiline notes, and text containing quotes. The converter writes quotation marks only when the destination format requires them and doubles quotes inside a quoted field. Open a small downloaded sample in the intended application before converting a business-critical export.
Preserve delimiters, line breaks, and Unicode text
CSV and TSV fields may contain their delimiter, quotation marks, or line breaks. The converter reads doubled quotation marks and writes quotes only where required, so a company such as “Café, North” remains one field when commas become tabs. Unicode text is preserved as UTF-8, and empty cells remain present even at the end of a row.
Every row must contain the same number of fields as the heading row. An incomplete or overfilled row stops with an error instead of shifting values into the wrong columns. CSV and TSV are both text-table formats, so keep the original file if another application depends on formula behavior, locale-specific dates, leading zeros, or a private null marker.
Limits, errors, privacy, and a practical final check
Empty input, unclosed or misplaced quotation marks, inconsistent row widths, duplicate headings, ambiguous aligned boundaries, and oversized tables stop with a visible message. The page limits rows, columns, total cells, and input length so an accidental paste cannot freeze the tab. For a very large export, use a streaming desktop or command-line workflow.
Reset clears the input, result, mapping notice, warning, and download. Before sharing a converted file, open it in the intended destination, count the rows, and inspect values containing punctuation, repeated spaces, quotes, or non-English text. A successful conversion confirms consistent field boundaries; it does not prove that a private import rule will accept every value.