Skip to the tool
RU

Find and remove duplicates

Choose the columns that define a duplicate — email, phone, order number or the whole row. Tablewise counts the copies across millions of rows in a second, shows the worst offenders and removes the extras, keeping the first row.

How it works

  1. Open a file

    Drop a CSV, Excel or Parquet file — or try the demo with a million sales rows. It opens in your browser, nothing is uploaded.

  2. Filter, group, chart

    Sort and filter columns, find duplicates, build a pivot table or a chart — every action takes a fraction of a second, even on millions of rows.

  3. Ask and save

    Ask a question in plain words, check the SQL, then download the result as CSV, Excel or Parquet.

Why Tablewise

Questions

Are “Ivan@Mail.ru” and “ivan@mail.ru ” duplicates?

Turn on “ignore case and spaces” — then letters are compared without case and spaces at the ends are cut off.

Is my file uploaded anywhere?

No. Tablewise opens the file inside your browser with DuckDB, a real analytical database compiled to WebAssembly. Rows, numbers and names never leave your device — you can switch off the internet after the page has loaded and keep working.

What does the “ask in words” feature send to the AI?

Only the column names and their types (for example “region — text, amount — number”) and your question. Not a single row of data is sent. The AI writes an SQL query, and the query runs on your computer.

How big can the file be?

Millions of rows. A 1 GB CSV with 10 million rows opens on a normal laptop; the limit is your device memory, not our servers. Excel files are read up to their own limit of 1 048 576 rows per sheet.

Is it free? Do I need an account?

Free, with no sign-up, no trial and no row limits. The work is done by your device, so it costs us almost nothing.

Which formats can I open and save?

Open: CSV, TSV, TXT, XLSX, XLS, ODS, Parquet, JSON. Save: CSV, Excel (XLSX, up to 1 048 576 rows) and Parquet — a compact format that keeps millions of rows small.

More table tools