Excel to CSV converter (XLSX to CSV)

Pick Monthly or Stations in the top bar and download again for theirs. A test opens the sample on every change to the site and compares each sheet’s CSV with the file we expect, byte for byte. Download the sample workbook (.xlsx, 6 KB) to try it in Excel too, or more sample CSV and Excel files.

What the CSV keeps, exactly

Values, not formulas

A CSV file holds values only. Each formula cell gives the result the workbook saved with it, as Excel saves it after every calculation. A workbook written by a program that never calculated its formulas has no saved results, and those cells come out empty.

Numbers as they are stored

A cell’s number format changes what Excel shows, not what it stores. The CSV writes the stored value with 15 significant digits, the precision Excel itself writes: a cell formatted to one decimal that shows 5.9 but holds 5.85 gives 5.85, and the tiny binary remainder of a calculation (5.8500000000000005) is dropped. For every digit, tick Full precision (17 digits) under “Do more”.

Dates and times

Excel stores a date as a number of days and shows it through its format. The CSV writes it in one unambiguous form, 2025-01-31, with the time when the column holds times (2025-01-31 09:30), whatever the format: Jan 2025, 31/01/2025 or 1-31-25 all come out the same way.

Percentages

A column formatted as percentages is read as percentages: a cell that holds 0.9 and shows 90 % is written 90. The column’s name stays as it is, so add “%” to it in Excel first if the reader of the CSV needs to know.

Text, TRUE/FALSE and empty cells

Text is written as it is, in double quotes when it holds a comma, a quote or a line break, as CSV readers expect. TRUE and FALSE stay TRUE and FALSE. An empty cell, or an error such as #DIV/0!, gives an empty field, and a note says how many errors there were. In a column of numbers, a cell that says the value is missing (n/a, NA, #N/A, null or a dash) gives an empty field too; any other text in it, such as a comment, is kept as it is.

One sheet at a time

A CSV file is one table, so each sheet is its own CSV. Title rows above the header are left out (a note says so), and so are empty rows and columns with neither a name nor values. Charts, colours, comments and merged-cell layout are not data and do not come along.

Old .xls and .xlsb files

MegaRows reads .xlsx and .xlsm workbooks, the format of every Excel since 2007. It does not read the older binary .xls format (Excel 97–2003) or the binary .xlsb workbook, and it says so plainly if you open one. Open the file in Excel or LibreOffice Calc, choose Save As and Excel Workbook (.xlsx), and convert the new file here. A password-protected workbook needs its password removed in Excel first.

Other ways to convert Excel to CSV

• In Excel: File › Save As › CSV UTF-8 (Comma delimited) saves the sheet that is open, as its cells are displayed, so a number formatted to one decimal keeps one decimal. Repeat for each sheet.
• In LibreOffice Calc: Save As › Text CSV, where the option “Save cell content as shown” decides between displayed and stored values.
• In Python: pandas reads a sheet with read_excel (with the openpyxl package installed) and writes it with to_csv.

The tool on this page needs none of them, and keeps the stored values. For a workbook going to MATLAB rather than to a CSV, convert it to a .mat file instead.

Private by design: nothing is uploaded

Workbooks often hold figures that must not leave the company. This tool reads your workbook and writes the CSV in your browser, so the file and its name never leave your device, and macros in an .xlsm file never run. It handles big sheets too, up to Excel’s 1,048,576 rows: the practical limit is your device’s memory.