Reconcile two CSV exports in your browser: auto-detected key matching, ignore-columns, value recoding, numeric tolerance, and a per-column breakdown of where the differences are. Reports as text, CSV, highlighted Excel or visual HTML diff. No upload.
Privacy: Everything runs locally in your browser. Your source file is not uploaded to JAD Apps servers.
Reconcile two CSV exports of the same data. Rows are matched on a key column — detected automatically, picked from a list, or typed — then compared cell by cell. You get the differences, the rows that exist on only one side, and a ranking of which columns account for the changes. Everything runs locally in your browser.
Free is enough for most one-off jobs. Pro raises the file and batch caps; Pro + Media unlocks GB-scale streaming and unlimited duration.
Larger files supported on Developer (5 GB CSV) and Enterprise (unlimited). All processing happens in your browser — files never reach a server.
Drop both CSV files
Add the baseline as file A and the comparison as file B. Use Swap A ↔ B if they land the wrong way round. Both are read in your browser; neither is uploaded.
Choose how rows are matched
Auto-detect scores every shared column by uniqueness and picks the best key. You can also select a column from the list, type a name, or fall back to sorted row-order alignment when the data has no identifier.
Silence the expected noise
Mark timestamp and audit columns as ignored, add value recoding rules where the two systems encode the same fact differently, and set a numeric tolerance if formatting differences like 1 versus 1.0 shouldn't count.
Pick your outputs
Choose any combination of the written report, long-format value differences, per-column breakdown, one-sided rows, the TRUE/FALSE mask, a highlighted Excel workbook, and a visual HTML diff.
Download the results
Take each artifact individually or all of them as a single zip. The summary always states which key was used and how it was chosen.
Every column the two files share is scored on how unique and how well-populated its values are in both files at once. A column must be at least 99.9% unique and 99% filled in both to qualify, with a small bonus for names like id, ref or code. If nothing qualifies, the tool says so and falls back to sorted row-order alignment rather than guessing.
Rows that duplicate each other exactly are dropped and counted. Rows that share a key but disagree on content are quarantined — pulled out of the match entirely and reported separately — because a fan-out join would silently multiply the differences.
Usually the two systems encode the same fact differently — one writes Y/N where the other writes 1/0. Add a value recoding rule for that column and the false differences disappear. If the column is a timestamp that changes on every export, mark it ignored instead.
Yes. A value that went from populated to empty, or empty to populated, is a real change and is reported. An empty string and a missing trailing cell are treated as the same thing.
Yes — type the names joined with a plus, for example account + period. Both columns must exist in both files, and their combination is used as the key.
No. Parsing, matching and every report — including the Excel workbook — are generated in your browser. Only anonymous usage counts are recorded, and only when you are signed in.
Privacy
Everything runs locally in your browser. Your source file is not uploaded to JAD Apps servers.
Merge multiple CSV files online by matching headers by column name. Free for 2 files, Pro for 3+ files. Browser-based, no upload.
Remove duplicate CSV rows by choosing a unique key column like Email, SKU, phone, or ID. Browser-based, no upload.
Run a CSV sanity check in your browser. Find row-width drift, type mismatches, duplicate headers, encoding issues, and empty cells.