Data workbench

CSV Compare & Diff Tool

Compare two CSV datasets by row position or stable key, inspect cell-level changes, and export a complete difference report locally.

Local by defaultYour input stays in this browser tab.
CSV Compare & Diff Tool Processed locally
Original CSV
Updated CSV
ResultAdd data or load the sample to begin.
StatusKeyChanged fieldsBeforeAfter

Basics

Understanding the format

A structured CSV diff compares records and fields rather than raw text. Key matching keeps the same record aligned when rows are reordered.

The result separates added, removed, changed, unchanged, and duplicate-key records.

What it does

What this tool does

Paste or open two files, choose a key such as id, email, or SKU, normalize harmless formatting differences, and filter a field-level audit table.

Key-aware matching

Match by a detected header or by row position when order matters.

Noise controls

Optionally ignore case, surrounding whitespace, and null-like placeholders.

Cell-level audit

See changed fields with before and after values plus source row numbers.

Complete export

Download the filtered or full difference report as CSV.

How to use

How to use it

  1. Add both versions

    Paste the original and updated CSV, open two files, or load the comparison sample.

  2. Choose matching

    Select a unique key; use row position only when both files share the same ordering.

  3. Compare

    Review duplicate-key warnings and counts for every result status.

  4. Filter and export

    Inspect only changed, added, or removed rows, then copy or download the report.

Errors & limits

Common errors and limits

Duplicate keys

Repeated identifiers are isolated instead of silently overwriting one another.

Different headers

Fields missing from one file appear as changes; align header names before comparing when they mean the same thing.

Formatting-only differences

Use the case, whitespace, and null controls deliberately; they change what counts as equal.

Review

A key must be stable and unique. Duplicate or missing keys are reported separately because matching them automatically could hide or invent changes.

FAQ

Frequently asked questions

Can reordered rows be compared?

Yes. Choose a stable key column and row order will not create false changes.

What key should I use?

Use a unique, stable field such as id, UUID, email, or SKU. The tool suggests likely keys.

Can I compare without a key?

Yes. Position mode compares row 1 with row 1, row 2 with row 2, and so on.

Are files sent to a server?

No. Both datasets and the diff stay in your browser.