A table copied from a spreadsheet often uses tabs, while another importer expects commas or receives a semicolon-separated regional export. The CSV and TSV converter changes those boundaries without interpreting amounts or identifiers. It works in both directions and returns tabular text for copying or export. Records and cells stay associated even when a quoted value contains a delimiter. This is not a workbook converter: it does not process Excel sheets, formatting or calculation formulas. It changes the representation of textual fields rather than the business information inside them.
For CSV to TSV, the source delimiter is detected from commas, semicolons, tabs and vertical bars using the first logical record. Output uses a tab between cells. For TSV to CSV, reading explicitly uses tabs and output always uses commas. This page offers no alternative destination CSV delimiter. If a receiving application requires semicolons, use a suitable export route rather than merely renaming the extension of a comma-separated file. File extensions describe formats but do not change field boundaries.
No record is treated specially as a header. Every record is converted in its original order, including files without headings and files with repeated heading names. Values remain strings, retaining leading zeros, spaces and decimal punctuation. Reading recognises quoted fields, doubled quotation marks and quoted line breaks. The serializer adds quotation marks when the destination delimiter requires them. That quoting convention supports complex TSV values, but some specialised TSV readers do not implement the same rules, so compatibility must be checked with the receiving software.
A leading UTF-8 BOM is removed, and newly written record separators use LF line endings. The precise use of quotation marks may therefore change without changing the cells. An unclosed quote prevents conversion and produces an error message. Equal record widths are not validated here; use CSV validation when that constraint matters. The import button accepts files up to 5 MiB and reads UTF-8 text. There is no Windows-1252 or UTF-16 encoding selector, nor a streaming mode intended for very large files.