OKfmt

Comparison of XLSX/ODS/CSV Spreadsheet Formats and Cross-Tool Exchange Guide

This article compares the feature support of three common spreadsheet formats, summarizes common pitfalls in cross-software data exchange, and helps office users select the correct format for data exchange.

Bijgewerkt 2026-08-11

Core positioning differences of the three formats

XLSX is a proprietary spreadsheet format launched by Microsoft Excel. It is now a closed-source format compatible with most office software, and is mainly used for local editing and saving in Excel.

ODS is an open spreadsheet standard launched by the Open Document Format Alliance. It is used as the native save format by LibreOffice and OpenOffice, has no copyright restrictions, and can be read by all mainstream office software.

CSV is a plain-text specification for storing tabular data. It only stores cell data and does not retain formatting or structure, and is positioned for plain data exchange between different tools.

Feature support comparison matrix

When converting spreadsheets, tools extract the cached calculation results of formulas for output by default, and do not retain the original formula text. This rule applies to both XLSX and ODS formats, and meets the requirements of most data exchange scenarios.

If you need to retain formulas for subsequent secondary editing, save the original file in the original editing software, and do not output and overwrite the original file with a new format, to avoid losing formula information.

FormatFormula retentionRetain cell stylesSupports multiple worksheets
XLSXSupportedSupportedSupported
ODSSupportedSupportedSupported
CSVNot supportedNot supportedNot supported

Common pitfalls of CSV format

When exchanging data with CSV, you should manually select UTF-8 encoding for saving, to avoid garbling caused by ANSI encoding. If the target tool cannot recognize semicolon-separated files, you can re-export a comma-separated version using the corresponding regional version of Excel.

CSV conversion tools can automatically recognize common delimiters and output CSV files with uniform UTF-8 encoding, which solves most format recognition abnormalities.

  • Excel saves CSV in the operating system's default ANSI encoding by default, rather than UTF-8 encoding, which causes Chinese character garbling when imported into LibreOffice or data analysis tools
  • Some regional versions of Excel for European regions use semicolons as CSV delimiters by default instead of commas, which causes column recognition errors when imported into other tools
  • Cells containing line breaks or delimiters that are not wrapped in quotes will be recognized by tools as line breaks or new columns, which breaks the table structure

Instructions for JSON conversion paths

JSON is a common data exchange format for program interfaces. Office users often need to convert Excel spreadsheets to JSON for interface docking, or convert JSON returned by interfaces into spreadsheets for analysis.

When converting XLSX/ODS/CSV to JSON, each row of table data is converted to a JSON object by default, with table headers used as object keys, and an array-format JSON result is output, which conforms to the specification requirements of most program interfaces.

Veelgestelde vragen

Which format should I choose for cross-tool spreadsheet exchange?

Choose ODS if you need to retain styles and formulas. ODS is an open standard that is compatible with all mainstream office software; choose UTF-8 encoded CSV if you only need to exchange data.

Why do formulas disappear from the spreadsheet after conversion?

Conversion extracts the cached calculation results of formulas for output by default, and does not retain the original formula text, which adapts to data reuse scenarios. Keep the original file if you need to retain formulas.

How to fix Chinese garbling when importing CSV into Excel?

You can use a CSV conversion tool to convert the encoding to GBK or UTF-8 with BOM, then re-import into Excel. Chinese characters will display normally and the garbling problem will be resolved.