Filter CSV by Column Value

1. 1

   Paste or upload CSV

   Paste CSV text or drop a .csv, .tsv, .txt or .psv file. Nothing is uploaded.

2. 2

   Check the columns

   Confirm the delimiter and whether the first row is a header. Each column shows the type it was read as.

3. 3

   Build your conditions

   Pick a column, choose an operator, and type a value. Add more conditions and choose whether rows must match all of them or any.

4. 4

   Read the live count

   The matched row count and the preview update as you type, so you can see the effect before exporting.

5. 5

   Export

   Download or copy the filtered CSV, choosing the delimiter, line ending and whether to add a BOM for Excel.

Build one or more conditions, see the matching rows immediately, and export only what you kept.

• Eighteen operators — Text, numeric, date and emptiness tests, including regular expressions and ranges.
• Several conditions at once — Combine conditions with match all or match any instead of filtering twice.
• Invert the result — Keep the rows that do not match, which is how you delete rows rather than select them.
• Type-aware columns — Each column is typed, so the operator menu and the comparison both fit the data.
• Live matched count — The count and the preview update as you type, before any export.
• Header-aware columns — The column list shows index and header label, and the header row can be overridden.
• Export options — Choose the delimiter, the line ending, and whether to add a BOM so Excel reads it correctly.

The menu adapts to the column you picked, so you are only offered comparisons that can actually answer something about that data.

Text operators

• Equals / Not equals — The whole cell matches, or does not. Use it for status codes and exact identifiers.
• Contains / Does not contain — The value appears anywhere in the cell. The usual choice for free-text notes and addresses.
• Starts with / Ends with — Prefix and suffix patterns — order references, SKU families, file extensions.
• Is one of — A comma-separated list, so three cities do not need three separate conditions.
• Matches a regular expression — For patterns the other operators cannot describe, such as a postcode shape.
• Is empty / Is not empty — Find the gaps in a column, or drop the rows that have them.

Number and date operators

• Greater than, less than, and their inclusive forms — Offered only where the column really holds numbers.
• Between — An inclusive range. The bounds can be typed in either order.
• Before, after, on — Date comparisons that understand ISO dates and the day-first and month-first spellings.

One condition answers a simple question. Combining them is where filtering starts to replace a spreadsheet.

• Match all — A row is kept only if every condition holds — shipped orders over 100 in London.
• Match any — A row is kept if at least one condition holds — anything cancelled or refunded.
• Invert — Flip the whole result. Describe the rows you want gone, then remove them in one step.

Each column is sampled and read as text, numbers or dates, and that decision changes both the operators offered and the comparison performed.

• Numbers are compared as numbers — On a numeric column, 1.0 and 1 and +1 are the same value, and 10 is greater than 9.
• Text is compared as text — On a text column a part number like 007 keeps its zeros and is not equal to 7.
• Currency, percentages and accounting negatives — $1,200.50, 12% and (45) are all understood as the numbers they represent.
• A stray value does not retype a column — One N/A in a column of quantities leaves it numeric, so greater than stays available.
• You can override the type — If a column of codes was read as numbers, set it back to text and the comparisons change with it.

Reading the file correctly comes before filtering it, and each of these settings is detected first and adjustable after.

• Header detection — A file has a header unless its first row looks like the data underneath, so a headerless file does not silently lose its first row.
• Delimiter detection — Each candidate is parsed and scored on how consistent the row widths are, so a comma inside a quoted address does not win.
• Quote character — Double quote, apostrophe, or none. Files quoted with apostrophes exist and reading them wrongly merges rows.
• Blank lines — A trailing newline is not a row. Interior blank lines can be kept or skipped, because sometimes they are padding.
• Encoding — UTF-8, UTF-16, Windows-1252 and ISO-8859-1, with any byte-order mark honoured automatically.

Both are per-condition, because the right answer differs between a customer name and a product code.

• Case sensitivity — Off by default, so London and london match. Turn it on for codes where case carries meaning.
• Trim whitespace — On by default, so a stray leading space does not hide a match. It applies to both sides of the comparison.
• Empty means empty — A cell of three spaces counts as empty, which is what a person reading the file would say.

A filtered CSV is only useful if the next program reads it the way you meant, which is mostly a question of three settings.

• Add BOM for Excel — Without it Excel reads a UTF-8 CSV as ANSI and accented characters arrive as mojibake.
• Line ending — LF or CRLF. Some older Windows tools require CRLF and will otherwise read the file as one line.
• Output delimiter — Keep the file's delimiter or change it, which is often easier than converting afterwards.
• Header row — Include or omit the header, depending on whether the result is going to a person or into an import.

The same conditions can be applied to a set of files, which is the usual shape of a monthly export split by region or by month.

• One set of conditions — Build the filter once and apply it to every file in the queue.
• One archive out — Every filtered file is delivered as a single ZIP rather than one download per file.
• Per-file counts — Each file reports its own matched and total row counts, so an empty result is visible rather than silent.

Filtering isolates the rows that matter for analysis, reporting, or the next import.

• Faster review — Focus on matching rows instead of scanning the full file.
• Cleaner exports — Share or import only the subset you need, without hidden rows travelling with it.
• Files a spreadsheet struggles with — A large export opens here without the wait, because only the rows you keep are written out.
• Nothing leaves the device — Customer records and financial data can be filtered without uploading them anywhere.

Filtering runs on your device, in the background, so the page stays responsive on large files.