CSV guide

CSV Delimiters: Comma, Semicolon, Tab or Pipe

A CSV delimiter is the character that separates fields in a record: usually a comma, but sometimes a semicolon, tab or pipe. If your file opens in one column or splits values incorrectly, check the actual separator before changing the file. Use the examples and local detector below to verify the columns and export with the separator you need.

Detect a CSV delimiter

Check your file here without uploading it. Compare the four separators.

Comma, semicolon, tab and pipe: CSV delimiter examples

These four records represent the same three values: Ada, London and the text 12,50. A separator outside quotes starts the next field; a separator inside a quoted value is part of that value.

SeparatorExample recordWhat to check
Comma ,Ada,London,"12,50"Quote a value when it contains a comma.
Semicolon ;Ada;London;12,50Often used by spreadsheet exports where decimals use commas.
Tab (TSV) TABAda → London → 12,50Each arrow represents one tab character, not spaces or a literal arrow.
Pipe |Ada|London|12,50Common in data exports; a pipe inside a value still needs quoting.

The .csv extension alone does not tell you the delimiter. Inspect several records, including one with quoted punctuation, and confirm the expected headers.

Step by step

  1. In the workbench below, choose your file or select Try sample. Review the detected delimiter, column count and header preview.
  2. If the columns look wrong, open CSV detection options, select comma, tab, semicolon or pipe, then choose Apply options and reparse. Detection is a suggestion, not proof of the intended schema.
  3. Check a quoted value and a decimal value. The supplied semicolon sample should have three columns and two data rows; 12,50 must stay in one amount cell.
  4. To change the CSV separator, open CSV export options and choose the output delimiter, then select Repaired CSV. Reopen that output and compare the row count and cell values before replacing the original.

Try the example on this page

Choose Try sample to load this guide’s synthetic data. Adjust the settings, inspect the result and export it here. You can also choose your own file for local processing.

CSV export options
Drop a file, paste data, load a CORS-enabled URL, or try the sample.
Rows0
Columns0
EncodingAuto
DelimiterAuto
Issue groups0

Open a data file to start

The parser runs off the main thread. Once loaded, you can search, sort, validate, repair and export without an account.

Original samples: semicolon.csv

Check expected outcomes and file checksums

Related CSV guides

Check the result

  • Use the column count and header preview to verify that every record follows the expected shape.
  • Inspect a value that contains the delimiter inside quotes, such as a postal address or product description.
  • If a file mixes delimiters between rows, treat it as malformed input and repair the source rather than trusting a single automatic choice.

Common pitfalls

  • Splitting each physical line with String.split breaks quoted commas and multiline fields.
  • A one-column file provides too little evidence for confident delimiter detection.
  • Decimal commas and semicolon-delimited regional exports are a common source of mistakes.

Fix the symptom, then check the separator

The whole CSV appears in one column

The importer may expect a different separator. Try the actual delimiter in the preview before saving. A genuine single-column file is also valid, so compare against the source system’s expected columns. See the Excel one-column troubleshooting guide for that specific workflow.

A decimal or an address becomes extra columns

In a comma-delimited file, write a value such as 12,50 or London, UK inside double quotes. In a semicolon-delimited file, the decimal comma is ordinary text. Do not globally replace separators: that can change cell contents as well as field boundaries.

Letters look wrong but the columns are correct

This is more likely an encoding issue than a delimiter issue. Check a known accented name with the CSV encoding guide. A UTF-8 BOM helps identify encoding; it does not change the separator.

Rows have different numbers of columns

Check for missing fields, unclosed quotes or mixed separators. A quoted field may contain a line break, so counting physical lines does not reliably count CSV records. Inspect validation issues before exporting; changing the delimiter cannot repair every malformed file.

How to choose a CSV delimiter in Excel

In supported desktop Excel versions, use Data → From Text/CSV, select the file, then choose the separator under Delimiter. Check the preview before loading. This lets you choose the separator for the import without changing the computer’s regional list separator.

For format conversion, use CSV to Excel to create an XLSX workbook. To rewrite a semicolon file as comma-separated text, follow convert semicolon CSV to comma.

Source: Microsoft’s text and CSV import instructions (checked September 11, 2026). Menu availability varies by Excel version.

CSV separator questions

Is a CSV separator the same as a delimiter?

In this context, yes: both mean the character separating fields. A quote character has a different job: it encloses values that contain separators, quotes or line breaks.

Can a CSV use a semicolon instead of a comma?

Many applications call semicolon-delimited exports CSV. Import them with semicolon selected and verify the columns. Use the delimiter required by the receiving application when exporting.

How can I detect the delimiter of a CSV file?

Load the file in the workbench above and inspect the detected separator and table. Automatic detection compares possible separators, but short or single-column files can be ambiguous. Override the choice and reparse when the preview does not match the expected data.

Does changing the delimiter also change decimal numbers?

No. This workbench preserves imported values as text. Exporting 12,50 from a semicolon file to a comma file produces a quoted field containing 12,50; it does not turn the value into 12.50.

Can this detector process a 1 GB CSV?

The embedded standard workbench retains up to 250,000 rows and disables export for incomplete data. For a large file, use the 1 GB tested streaming viewer and select its delimiter manually. Its CSV output uses commas.

Privacy boundary

Files selected in the workbench are processed locally. A public URL is requested directly by the browser and must permit cross-origin access. No server proxy is used by the MVP.