CSV Split, Merge and Remove Duplicates
Break a large CSV into smaller files, combine several CSVs into one, or drop rows that share the same key columns. Quoted fields with line breaks are kept intact. To browse or sort a file as a table, use the CSV Viewer; for plain archives see the ZIP Tool.
🔒 Runs in your browser · Your data is never sent to a server
Choose a CSV, TSV or text file, or drag & drop it here
Choose two or more CSV files, or drag & drop them here. They are merged in the order selected
Files to merge
Choose a CSV, TSV or text file, or drag & drop it here
How to use
- Under "Common settings", choose the encoding, delimiter and whether the first row is a header.
- Pick the "Split", "Merge" or "Remove duplicates" tab and choose your CSV file(s) (drag & drop works too).
- For splitting, set a row count or number of files; for duplicates, choose key columns and which row to keep; for merging, select several files.
- Press "Split and download ZIP", "Merge and download" or "Remove duplicates and download" to save the result.
Notes
- Every split file gets the same header row (when "First row is the header" is on). The parts are saved together in one ZIP and named like original_01.csv.
- Merging keeps the first file's header and drops the header rows of the others. With "Check that headers match" on, any file whose column names, order or count differ by even one character stops the merge with an error. With it off, the later header rows stay in as data rows.
- Duplicate removal keeps the remaining rows in their original order. Values are compared as exact text, so 1 and 01 are different. The header row is never removed.
- Output is always UTF-8 with LF line endings, quoting only the cells that need it. Quoting style and CRLF line endings of the original may change. Completely blank lines are dropped.
- The whole file is loaded into browser memory, so files up to 100 MB are accepted (a warning appears above 20 MB). Everything runs in your browser and files are never uploaded.
Glossary
BOM
A marker at the start of a file. With it, Excel opens a UTF-8 CSV without garbling non-ASCII text.
Key column
A column compared to decide whether rows are duplicates. Using an "ID" column as the key makes rows with the same ID duplicates.
FAQ
What is the difference between rows per file and number of files?
Rows per file cuts the data every N rows, such as 1,000, and only the last file holds the remainder. Number of files picks how many parts you want and spreads the rows as evenly as possible. Either way the original row order is kept and all parts are saved in a single ZIP.
When does merging fail with an error?
With "Check that headers match" on, merging stops if any file differs from the first in column names, column order or column count, and it names the file. Merging CSVs whose columns are arranged differently would shift your data, so align the columns first, for example in the CSV Viewer.
Which columns should I use as the key for removing duplicates?
Choose the column(s) that identify the same person or item, such as an ID or email address. If you choose none, only rows that are identical across every column count as duplicates. With several columns, all of them must match. You can keep the first or last occurrence, and the remaining rows stay in their original order.
The file looks garbled when opened in Excel.
Turn on "Add a BOM to the output" so Excel recognizes the file as UTF-8. When you load a Shift_JIS CSV, pick Shift_JIS under "Input encoding" instead. The output is always UTF-8.
Can it split a CSV that has line breaks inside cells?
Yes. Line breaks and commas inside quoted cells are treated as part of that cell, so a split never cuts through a cell and rows do not shift.
Share this page