Deduplicate a CSV by complete rows or selected key columns while controlling whitespace, case, and empty values locally.
The first row is treated as column names.
Rows are duplicates when all selected comparison columns match.
Parse the CSV to choose columns.
The original row order and column order are preserved.
Two CSV rows are not always duplicates just because every field matches. Customer exports may need comparison by email, inventory files by SKU, and event files by a combination of user ID and timestamp. This tool lets you choose the columns that define a duplicate while preserving the first matching row and the original file order.
Ada@example.com and ada@example.com.Normalization is used only for comparison. The values written to the output remain exactly as they appeared in the first retained row.
If email is the selected key, these rows match when case and whitespace are ignored:
email,name
ada@example.com,Ada Lovelace
ADA@example.com, Ada Lovelace
The output keeps the first row. Select both email and name if both fields must match before a row is removed.
Parsing and deduplication happen entirely in browser memory. Your CSV is not uploaded to MiniUtil. Files are capped at 20 MB to avoid locking the tab; much larger datasets are better handled with streaming database or command-line tools.
The parser preserves quoted delimiters, embedded line breaks, and escaped quotes. It does not infer numeric types because values such as ZIP codes and account numbers may require leading zeroes.
Use the CSV Delimiter and Encoding Fixer before deduplication when a file opens in the wrong columns, the CSV Viewer to inspect the cleaned table, or CSV to JSON and the JSONL Converter for structured-data exports.
Share this utility
Send the tool, not any text or files you entered. Input is never added to the share URL.
Continue the workflow
Step 3 of 5. Keep the context and finish the related checks.
Keep working