CSV Cell Diff
Paste two CSVs to align them column-by-column and see exactly which cells changed — with optional key-column row matching, all in your browser.
CSV Cell Diff
Compare two CSV files column-by-column and see exactly which individual cells changed — not just which lines differ. Unlike a plain text diff, this tool understands column structure: it lines columns up by header name (so reordered columns still match) and can pair rows by a key column (so reordered rows still match), then reports every changed cell with its old and new value, plus which rows and whole columns were added or removed. Everything runs locally in WebAssembly — your data never leaves the browser.
Worked example
Original CSV
id,name,price
1,Apple,10
2,Banana,20
3,Cherry,30
Updated CSV
id,name,price
1,Apple,12
2,Banana,20
4,Date,15
With Key column(s) = id and Output format = table:
3 columns compared
1 rows changed · 1 rows added · 1 rows removed · 1 rows unchanged · 1 cells changed
~ [1] price: "10" → "12"
- [3] id=3, name=Cherry, price=30
+ [4] id=4, name=Date, price=15
Row 1's price cell changed 10 → 12, row 4 is new, and row 3 was removed. Row 2 is
unchanged. Switch Output format to json for a structured report (with per-cell
old/new and summary counts) or csv for a flat row_key,status,column,old,new
change-log you can pipe into another tool.
The same example through the gizza CLI:
gizza tool csv-cell-diff left=$'id,name,price\n1,Apple,10\n2,Banana,20\n3,Cherry,30' right=$'id,name,price\n1,Apple,12\n2,Banana,20\n4,Date,15' key=id format=table
Inputs
- Original CSV / Updated CSV — the two versions to compare.
- Key column(s) — one or more column names (or 1-based indices), comma-separated, used to pair rows so reordering doesn't register as a change. Leave empty to pair rows by position (row 1 vs row 1, and so on).
- Delimiter — comma, tab, semicolon, or pipe (applied to both CSVs).
- First row is a header — on by default; columns are then aligned by name. Turn it off
to align columns by position (
col1,col2, …). - Ignore case / Ignore whitespace — fold case or collapse whitespace when comparing, while the output still shows the original text.
- Output format — a readable table report, a structured JSON report, or a flat CSV change-log.
Limits & notes
- Both inputs are pasted text; there is no file upload. Very large CSVs are limited by browser memory.
- Comparison is per-cell exact string match (after any case/whitespace folding); numbers
like
10and10.0are treated as different text. - With a key column, duplicate key values are paired in order of appearance; any leftover rows on either side are reported as added or removed.
- A whole column that exists on only one side is listed as added/removed and its cells are not compared against the other file.
FAQ
How is this different from a plain text diff?
A text diff compares whole lines, so moving a row or reordering columns shows up as a big
change. This tool parses the CSV structure: it aligns columns by header name and (with a
key column) matches rows by identity, then reports the specific cells that differ. The
result points at row → column: old → new, not just "this line changed".
What does the key column do, and can I use more than one?
The key column(s) identify each row so the tool can match the same record across both
files even if the rows were reordered. Give a single column like id, or several
comma-separated columns like first,last for a composite key. Leave it empty to compare
rows positionally (row 1 vs row 1). Columns can be referenced by header name or by 1-based
index.
My two files have the columns in a different order — is that a problem?
No. When the first row is treated as a header, columns are aligned by name, so a file
with price,id,name compares correctly against one with id,name,price. Only columns
that exist in both files are compared cell-by-cell; a column present on just one side is
reported as added or removed.
When should I pick the JSON or CSV output instead of the table?
Use table for a quick, readable review. Use json when a script needs the
structured result — it includes the common/added/removed columns, summary counts, and a
per-row list of changed cells with old and new values. Use csv for a flat
row_key,status,column,old,new change-log that you can open in a spreadsheet or feed into
another CSV tool.
Does my data get uploaded anywhere?
No. The comparison runs entirely in your browser via WebAssembly. Nothing is sent to a server, so it is safe to diff private or sensitive exports.
Developer & Automation Access
Run it from the terminal
Same engine as this page, headless — via the gizza CLI:
gizza tool csv-cell-diff "id,name,price
1,Apple,10
2,Banana,20" 'right=id,name,price
1,Apple,12
2,Banana,20'New to the CLI? Get gizza →
Open it by URL
Pre-fill and auto-run this tool with query parameters — the names match the API/CLI:
https://gizza.ai/tools/csv-cell-diff/?left=id%2Cname%2Cprice%0A1%2CApple%2C10%0A2%2CBanana%2C20&right=id%2Cname%2Cprice%0A1%2CApple%2C12%0A2%2CBanana%2C20&key=id&delimiter=comma&header=true&ignore_case=true&ignore_whitespace=true&format=tableMachine-readable descriptor: tool.json — title + parameters JSON Schema for agents.