Excel to TSV Converter

What the TSV export writes

One line per worksheet row, cells separated by a single tab character, rows separated by a line feed, and no trailing newline. Every row is padded out to the width of the widest row first, so each line carries the same number of tabs and a parser that counts columns does not trip over a short row at the end of the sheet.

Quoting is kept to a minimum. A cell is wrapped in double quotes only when it contains a tab, a double quote, a carriage return or a line feed, and a double quote inside such a cell is written twice. Everything else goes out bare, which is what keeps a TSV readable in a terminal and diffable in Git.

Region	Note	Revenue
North	Q1 revised	42000
South	"He said ""ship it"""	36500.5
East		

How each kind of cell exports

• Numbers - written as stored, with floating-point noise trimmed to twelve significant digits, so a subtraction Excel displays as 57.81 is not exported as 57.809999999999945.
• Dates and times - their underlying serial numbers. A cell showing 15 March 2023 exports as 45000, and 09:30 as 0.395833333333 of a day. Number formats are display rules and are not part of the value.
• Currency and percentages - the bare number: 1234.5 and 0.15.
• Formulas - the value the workbook last calculated, never the formula text.
• Booleans - true and false, lower case.
• Error cells - #N/A, #DIV/0! and the rest export as empty fields.
• Merged ranges - the value appears once, in the top-left cell; the cells it covered export empty.

Encoding, and the byte order mark

The download is UTF-8. A three-byte byte order mark is added when - and only when - the text contains a non-ASCII character. That mark is what makes Excel on Windows read a double-clicked .tsv as UTF-8 rather than as the system code page, so a name with an accent survives the round trip. A purely ASCII export is written without one, so the loaders and scripts that have never had to strip a mark do not suddenly have to start.

The other three formats in the Output format dropdown - Markdown, LaTeX and YAML - never get a mark, because in those a stray three-byte prefix is noise or a syntax error rather than a hint.

Edge cases

• The header checkbox does not affect TSV. First row as headers changes Markdown, LaTeX and YAML output. TSV always writes every row, header row included, because a TSV has no way to mark one row as different.
• One sheet at a time. A multi-sheet workbook gets a Sheet dropdown; each export covers the sheet you pick. Dropping several workbooks at once switches to batch mode, which exports the first sheet of each and returns a ZIP.
• The used range is exported whole. Blank rows above the table and blank columns to its left become empty fields, because the workbook declares them as part of the range.
• Hidden rows and columns still export. Hiding is a view setting, not a change to the data.
• Line breaks inside a cell survive. They trigger quoting, so the record stays one logical row for a parser that understands quoted fields - and looks like two lines to one that does not.
• Formatting does not cross. Fonts, fills, borders, column widths, conditional formatting, charts and images have no representation in tab-separated text.

When to use a different tool

• Commas rather than tabs, for a tool that expects RFC 4180: Excel to CSV.
• Going the other way, tab-separated text back into a workbook: TSV to Excel.
• A table for documentation or a repository README: Excel to Markdown; for a paper, Excel to LaTeX; for a config file or fixture, Excel to YAML. All four formats come from the same Output format dropdown on this page.
• Loading straight into a database: Excel to SQL writes INSERT statements instead.
• Every sheet as its own file: Excel Splitter first, then export the pieces.

Questions specific to TSV export

Why does TSV exist when CSV is everywhere?

Because tabs almost never appear inside real data, while commas appear constantly - in addresses, in product names, in decimal numbers across most of Europe. A TSV therefore needs quoting far less often, which makes it easier to read by eye and to split with cut or awk. It is also the format your clipboard already uses: copy a range in Excel and what lands on the clipboard is tab-separated text.

Can I paste the output straight into a spreadsheet?

Yes - that is the shortest route between two spreadsheets. Copy the output and paste into Excel, Google Sheets or LibreOffice Calc and the tabs become columns without an import dialog. Quoted cells containing line breaks are the one case where paste behaves differently between applications.

Should I use .tsv or .txt as the extension?

The download is named .tsv. Windows does not associate that extension with Excel by default, so double-clicking may not open it; File, Open in Excel does, and so does renaming it to .txt, which routes it through the import wizard where you can set each column's type.