CSV De-duplication Tool — Remove Duplicate Rows by Key Columns

Select CSV

Or paste:

Key Columns & Basic Settings Delimiter: First row is header
Key columns (comma-separated; multiple allowed): * Numbers are treated as 1-based column indexes from the left (1,2,3…); text is matched to header names. Mixing is OK.

When duplicates are found:

Trim leading/trailing spaces in keys Case-insensitive keys Collapse consecutive spaces to one

Output Settings Output delimiter: Include header in output

Table of contents

How to use (3 quick steps)
What you get (output)
Notes & caveats

How to use (3 quick steps)

Load a CSV

Upload a file or paste CSV text into the box.
If your file has a header row, leave “First row is header” checked.
Choose the correct delimiter (Comma / Tab / Semicolon / Pipe).

Set key columns & options

Key columns: specify by 1-based index (e.g., 2 or 2,5,7) or by header name (e.g., ID or ID,Date). Mixing is OK.
When duplicates are found: keep first occurrence (drop later rows) or keep last (overwrite earlier rows).
Key normalization (optional): trim spaces, case-insensitive, collapse repeated spaces to one.

Run & download

Click “Remove Duplicates” to see a preview (up to 200 rows).
Click “Download CSV” to save the de-duplicated result.

What you get (output)

A cleaned CSV with duplicate rows removed according to your key columns and the keep rule.
Header handling: if “Include header in output” is checked and your input had a header, it is preserved at the top.
Delimiter: output uses the delimiter you select (Comma / Tab / Semicolon / Pipe).
Encoding: output is UTF-8 (safe for modern tools; for Excel, import as UTF-8 to avoid garbling).

[Completely Free] Utility Tools & Work Support Tools

You can use practical tools like CSV formatting, PDF conversion, and ZIP renaming entirely in your browser, all for free. Each tool page clearly explains “How to use it”, “What the output looks like”, and “Important notes & caveats”, so even first-time users can start without confusion.

View all free tools

Notes & caveats

Column addressing: numeric indexes are 1-based from the left (1,2,3…); header names must match exactly (before normalization rules are applied).
Normalization scope: trimming, case-folding, and whitespace collapsing apply to key comparison only; non-key columns remain unchanged.
Empty keys: rows whose key fields are empty (after normalization) count as the same key; only one such row will remain per the keep rule.
Row shape: if rows have differing column counts, the tool pads missing cells with blanks so the table stays rectangular.
File size: default upload limit is 5 MB (configurable via max_mb).
Access control: by default the tool is limited to logged-in users (override with require_login=0 to publish).
Encodings: loader auto-detects UTF-8 / SJIS-win (CP932) / EUC-JP / JIS and converts to UTF-8. If characters look wrong, verify delimiter and source encoding.
Excel tips: import the downloaded CSV via “Data → From Text/CSV” and specify UTF-8.