Filter CSV rows by one rule

Updated 28 September 2026

To filter a CSV, tell the tool which column to check, how to compare it and what to compare it with. Rows that pass the test go into a new CSV, and rows that fail are left out.

It is a quick way to cut a huge export down to size without loading it into a spreadsheet, and it repeats cleanly each time a fresh export arrives.

Building the rule

Every rule names a column, picks one of four tests and supplies a value:

• Contains: the value can sit anywhere in the cell, so Portland finds that city inside a full street address.
• Equals: the cell has to match the value as a whole.
• Greater than and less than: numeric tests, for totals over 250 or stock under 10.

Case makes no difference to the text tests. Denver, denver and DENVER count as the same word for contains and for equals, and equals also forgives stray spaces before or after the content of a cell.

The numeric tests need plain numbers that use a dot as the decimal point, like 89.90. A cell holding 1.250,50 or $99 is not a plain number, so its row stays out of the result.

Type the column name the way the header shows it, with the same capital letters and spaces. When the name is wrong, the error message lists the columns the file has.

What goes in and what comes out

Input files need a header row. We detect commas, semicolons, tabs and pipes automatically, and read UTF-8 correctly with or without a byte order mark. The result keeps all of the columns and the original header, and it uses the same separator as the input.

Stacking rules and running batches

A single run checks a single rule. Two conditions mean two passes: keep the rows for one state, then send the smaller file through again with the amount rule. To make that repeatable, set up the two filters as consecutive steps in the pipeline builder and point it at a whole folder. Pipelines can be tried for free on up to 3 files.

Plain batches work too. Without an account you can drop up to 5 CSV files together, each up to 50 MB, within 20 server jobs a day. A free account raises that to 10 files per batch and 50 jobs a day. Pro takes 500 files of up to 500 MB and has no daily allowance. Collect the results one at a time or in a single ZIP.

The filtering happens on our servers in Germany, not in your browser. Your files travel over an encrypted connection, and we remove them automatically within two hours.

How to keep only the rows you need

1. Drop one or more .csv files onto the page.
2. Enter the column name as the header writes it.
3. Choose a test and type the value to compare against.
4. Press Filter and save the smaller CSV.