Extract columns from CSV

Type the column numbers in the Columns field. Columns are numbered from 1, left to right.

The table under the result shows the first three rows of every column with its number, so you can see at a glance which number is the email, the name or the price. If you ask for a column that doesn't exist, the output keeps an empty field and the note tells you.

Auto-detect looks at the first rows and picks the separator that appears consistently: tab, semicolon, comma or pipe, falling back to spaces. You can also choose it yourself, or enter a custom one like ::.

Data copied from Excel or Google Sheets is tab-separated, so it's detected automatically.

Handle quoted fields (on by default) follows CSV rules: a field in double quotes can contain the separator, line breaks and doubled quotes (""). So "Smith, Anna" stays one field instead of splitting at the comma.

For the output you can keep the same separator, or switch to comma, semicolon, tab (best for pasting into a spreadsheet), space, pipe or a custom string such as -. Quotes in output adds quotes only where needed (fields containing the separator, quotes or line breaks), around every field, or never.

• First row is a header, skip it. Leaves out the header row from the output.
• Trim spaces from fields. Removes spaces around each value, common in hand-edited files.
• Skip empty lines. Ignores blank rows.

Open .csv, .tsv and .txt files directly with Open file, and download the result as a text file. The file is read by your browser and never uploaded, which matters for customer lists and other personal data.

• Semicolon CSV. Spreadsheet programs in many European locales save CSV files with semicolons, because the comma is their decimal mark. Auto-detect handles this; if you send the result to a US system, switch the output separator to comma.
• Excel and leading zeros. ZIP codes like 02134 lose their leading zero if Excel opens the file as numbers. Extracting the column here keeps the text exactly as it is.
• Byte order mark. Some exports start with an invisible BOM character, which can stick to the first header name. If the first column name looks odd, clean it with invisible character.
• Ragged rows. If some rows have fewer fields, the missing columns stay empty in the output rather than shifting data.

Input:

Name,City,Email,Notes
"Smith, Anna",Boston,anna@example.com,"She wrote: ""on my way"""
Bob Jones,Denver,bob@example.com,

Columns 3,1 with header skipped and output separator comma:

anna@example.com,"Smith, Anna"
bob@example.com,Bob Jones

The name with a comma is quoted again in the output so the CSV stays valid.

Useful next steps: remove duplicate lines to dedupe the extracted emails, comma separator to turn a single column into one line, or merge lines to stitch columns from different sources back together.