Format Comparisons
Straight answers to the "which one should I use" questions, with the trade-offs laid out side by side. Each page ends with a recommendation, not a shrug.
How to choose a delimiter in one question
Nearly every format argument reduces to the same test: could a value in your data contain the separator character? If yes, you need either quoting or a different separator.
- Names, addresses, descriptions — contain commas constantly. Use tabs.
- European numbers —
1,50means one and a half. Use semicolons. - Log lines and legacy feeds — pipes, because text almost never has one.
- Numeric IDs and codes — anything works. Use commas for compatibility.
Comparisons in this section
| Comparison | Short answer |
|---|---|
| CSV vs TSV | Use TSV when your data contains commas — names, addresses, descriptions, anything written by a human. Use CSV when it does not,… |
| Comma vs Semicolon | The decimal separator decides it. In locales where a comma means a decimal point (Germany, France, Spain, Netherlands, Brazil),… |
| WPM vs CPM | WPM is the standard everywhere it matters — job assessments, certificates and published averages all use it. CPM is simply WPM… |
Frequently Asked Questions
Which delimiter should I use for messy data?
The tab character. Real text contains commas and semicolons constantly but almost never contains a tab, so tab-separated files stay valid without any quoting.
Is CSV or TSV better?
Neither is universally better. TSV is safer when values may contain commas, such as names and addresses. CSV has wider support in older software. The CSV vs TSV comparison on Delimiter.site sets out both cases.
Why do some CSV files use semicolons?
In locales where the comma is the decimal separator, such as Germany, France and Spain, spreadsheets export CSV with semicolons so that a number like 1,50 stays unambiguous.