Drop Duplicate Rows

Data > Drop duplicate rows... (Ctrl+Shift+H) removes repeated rows from the active table in one step.

How it works

Tick the columns that make up the key. Two rows count as duplicates when all their checked columns are equal. With every column ticked (the default) only exact whole-row repeats are removed; tick just one column to collapse rows that share that value.

Choose whether to keep the first or keep the last occurrence of each key, then press Apply. The rest are removed as a single undoable step (Ctrl+Z restores them all), and the status bar reports how many rows went.

Values are compared as text, so 1 (integer) and 1.0 (float) are not treated as the same. The operation respects read-only mode.

Command line and assistant

The same engine is available as octa --dedupe (see the --dedupe reference) and as the drop_duplicates MCP / assistant tool. To report duplicates instead of removing them, see Find Near-Duplicates or the search bar's Find duplicates highlight.