Save a CSV Without Losing Accented or Non-Latin Text

To create a CSV with UTF-8 encoding, select the application’s explicit UTF-8 option when saving or exporting.

Quick reference: UTF-8 export options by application

Use the workflow that matches your application and operating system. The documented procedures distinguish between Notepad and Excel on Windows and Numbers on macOS.

Application	Operating system	Option to select
Notepad	Windows	Save As → All Files → UTF-8; use a filename ending in .csv
Microsoft Excel	Windows	CSV UTF-8 (Comma delimited) (.csv)
Apple Numbers	macOS	File → Export To → CSV → Advanced Options → Unicode (UTF-8)

These option names follow the documented UTF-8 save and export procedures for Notepad, Excel and Numbers.

The Excel procedure is specifically for Excel on Windows. Do not assume that its menus or format choices are identical in Excel for macOS. The Numbers instructions are a separate workflow for Apple Numbers on macOS.

What UTF-8 encoding means for a CSV file

CSV and UTF-8 describe different properties of a file:

• UTF-8 is one possible character encoding.

The registered media type for CSV is text/csv, and its optional charset parameter can identify the character encoding. The relevant RFC context describes US-ASCII as the default while permitting other encodings through that parameter; it does not require every CSV file to use UTF-8. RFC 7111 is an Informational document primarily about URI fragment identifiers for CSV resources, although it also contains the relevant media-type registration context.

Individual importers can impose narrower requirements. MeisterTask, for example, requires imported CSV files to use UTF-8. That is an application-specific import requirement, not a universal rule for every CSV file or importer.

Likewise, selecting a comma as the delimiter controls how fields are separated; it does not establish that the file is UTF-8.

How to save a UTF-8 CSV in Notepad on Windows

Before saving, inspect representative accented and non-Latin characters in Notepad. Continue only if they display correctly. If the text is already garbled or contains unexpected replacement characters, stop and return to a known-good source rather than treating Save As UTF-8 as a repair.

To save the correctly displayed text as a UTF-8 CSV:

1. Open the CSV in Notepad.
2. Choose File → Save As.
3. Set Save as type to All Files.
4. Set Encoding to UTF-8.
5. Enter a filename ending in .csv, such as multilingual-data.csv.
6. Select Save.

Both selections matter in this workflow. The UTF-8 setting controls how Notepad encodes the saved text, while the .csv extension gives the output the expected filename type. Renaming the original file without selecting UTF-8 does not convert its encoding.

Save the result as a new copy rather than replacing your only source file. Keeping the original untouched gives you a clean version to return to if the exported copy does not open or import as expected.

How to save a UTF-8 CSV in Excel for Windows

In Microsoft Excel for Windows:

1. Open the file in Excel.
2. Choose File → Save As.
3. Open the file-format list.
4. Select CSV UTF-8 (Comma delimited) (.csv).
5. Choose a filename and save the new file.

Select the complete format name, including UTF-8. Do not treat an unspecified CSV option as equivalent merely because it also creates a filename ending in .csv.

These sourced steps apply to Excel on Windows. They do not establish the corresponding menu names or procedure for Excel on macOS.

How to export a UTF-8 CSV from Numbers on macOS

In Apple Numbers on macOS:

1. Open the spreadsheet in Numbers.
2. Choose File → Export To → CSV.
3. Open Advanced Options.
4. Set Text Encoding to Unicode (UTF-8).
5. Continue to the next export screen.
6. Choose a filename and location, then save the export.

The UTF-8 setting is inside the CSV export’s advanced options. This is a Numbers export workflow and is distinct from the Save As procedure documented for Excel on Windows.

Run a small pre-publication character check

Before sharing the complete dataset, test values that make character damage easy to notice:

label,value
plain_ascii,Cafe
accented,Café
city,München
non_latin,東京

Use the sample as part of a cautious export check:

1. Retain an untouched copy of the source dataset.
2. Export a separate CSV using the application’s documented UTF-8 option.
3. Reopen the exported copy.
4. Inspect the headers and representative rows containing plain ASCII, accented characters and non-Latin text.
5. Confirm that the displayed values still match the source before sharing or publishing.

Readable characters are a useful visual check, but they do not conclusively validate the file’s encoding. One application may interpret the file as intended while another importer handles it differently.

If the source text is already corrupted, saving it again as UTF-8 cannot be assumed to reconstruct the original characters. Return to an intact source copy instead of repeatedly resaving the damaged version.

Encoding also does not correct unrelated CSV problems such as incorrect delimiters, broken quoting or malformed rows. Investigate those separately if the text looks right but fields or records still import incorrectly.

Prepare the corrected CSV for public sharing

Keep the corrected UTF-8 export as a separate publication copy instead of overwriting the source dataset. Review the column headers, accented names, non-Latin values and representative records from different parts of the file before publication.

TablePage turns spreadsheets into public interactive data pages that readers can explore. Because a published dataset is public, remove private credentials, confidential records and personal data that should not be disclosed before uploading it to TablePage.

Publish only after the representative values in the corrected copy match the source. Choose the application’s explicit UTF-8 option, preserve the original dataset and inspect a small multilingual sample. The .csv extension alone does not establish encoding, and UTF-8 is an application choice rather than a universal CSV mandate.