Fix Excel-Damaged CSV Files in Seconds
Excel turned your barcodes into 1.23E+12, deleted the zeros in your ZIP codes, or filled your file with é garbage? Paste or drop the file below — get a clean CSV back.
Before you drop anything: press F12 and open the Network tab. Repair a file and watch — no request appears, because there is nowhere for one to go. The full 30-second check →
…or paste the data straight into the box below
Reading your file…
Nothing is being uploaded — this is running on your computer.
No CSV to hand? Download a deliberately broken one and drop it back in: collapsed barcodes, missing ZIP zeros or garbled names — or see all six samples.
Excel-safe export wraps long numbers and zero-led codes as ="00501", so Excel reads them as text and cannot re-damage them when the file is reopened. Leave it off if the file goes straight into another system, which may show the = literally.
🔒 Try it: switch off your Wi-Fi — the repair still runs, because your file never leaves this tab. We can't read, store, or leak what we never receive. Prove it in 30 seconds →
How it works — and why your data never leaves your device
1️⃣ Drop or paste your CSV
Drag the file onto the box above, choose it from your computer, or paste the rows straight in. The file is read by your browser’s own FileReader — it is not sent anywhere.
2️⃣ Pick the repairs
The tool analyses the data on your machine and lists what it found — scientific notation, missing zeros, broken characters. You tick the fixes you want; you stay in control of every change.
3️⃣ Download the clean file
Get a repaired CSV back instantly, with an Excel-safe option so the fixes survive the next time the file is opened. A live preview highlights exactly which cells changed.
🔒 Is my data safe? Yes. The repair engine is JavaScript running in your browser tab — switch off your Wi-Fi after the page loads and it keeps working. Your file is never uploaded, stored, logged, or seen by anyone, and there is no account to create.
Free CSV tools, all in one place
🔢 Scientific notation → full numbers
Excel collapses long UPC, EAN and account numbers into 8.85909E+11. We expand them back to the complete digit string using exact string math — no precision loss.
0️⃣ Restore leading zeros
ZIP code 00501 became 501? Product code 0071589 became 71589? Pick the column and the width — zeros come back instantly.
🔤 Fix broken characters
é instead of é, “ instead of quotes? That's an encoding accident (mojibake). We reverse it — even double-encoded files.
🔄 Convert CSV ⇄ TSV
Switch a file between comma, tab, semicolon and pipe delimiters for a fussy importer — with a preview and an Excel-safe option so your codes survive the trip.
Open the converter →🧹 Clean a CSV
Remove duplicate rows, drop blank rows, trim stray spaces, and delete empty columns — in one pass, with a before/after preview. Your header stays put.
Open the cleaner →🔎 Check CSV encoding
Not sure if your file is UTF-8, has a hidden BOM, or is legacy ANSI? Drop it in and get a plain-English report — plus a mojibake warning and the fix. Nothing uploaded.
Open the checker →🔗 Merge CSV files
Combine many CSVs into one — monthly exports into a yearly file — stacked and aligned by column name, with an optional source-file column. All in your browser.
Open the merger →✂️ Split a large CSV
Too many rows for Excel to open? Split a big file by row count or into a set number of parts — header repeated in each — and download them all as a zip.
Open the splitter →🔗 Join CSVs by a column
Match two files on a shared key — a customer list to its orders by ID — the VLOOKUP job without the formula. Left or inner join, in your browser.
Open the joiner →🛍️ Fix a Shopify product CSV
Shopify import failing, or barcodes, SKUs and titles arriving broken? Repairs all three Excel-caused problems at once — scientific-notation barcodes, dropped SKU zeros and garbled titles — ready to re-import.
Open the tool →📅 Fix date formats
Excel read 04/11/2023 in the wrong order and saved? Convert whole date columns between DD/MM, MM/DD and ISO — you state the source order, ambiguous dates are never guessed.
✅ Validate a CSV
Will an importer reject this file? Checks delimiter, unbalanced quotes, ragged rows and duplicate headers against RFC 4180 — nothing flagged by shape alone.
Open the validator →🩺 CSV Doctor
Not sure what’s wrong, or how many things are wrong at once? Run every check above in a single pass, then select and apply the safe fixes right there.
Open CSV Doctor →🔍 CSV Compare
What actually changed between two versions of a file? Compare by row position or by a key column like SKU or ID — duplicate keys are flagged, never guessed at.
Open CSV Compare →🧰 All tools & guides
Browse every free tool on one page, or read the guide that explains your exact problem and how to stop it recurring.
See all tools →Why did Excel ruin my CSV in the first place?
A CSV file is plain text — it has no way to say “this column is text, not a number.” So when Excel opens one, it guesses the type of every cell. Three guesses go wrong constantly in real business files:
- Long digit strings become scientific notation. Any number with 12 or more digits — UPC and EAN barcodes, container numbers, account IDs — gets displayed as
1.23457E+12. Save the file, and the notation replaces your real data. Marketplace and feed uploads (Amazon, Shopify, Google Merchant) then reject the file. - Leading zeros are deleted. To Excel,
00501is just the number 501. US Northeast ZIP codes, phone numbers, GTIN-8 codes and employee IDs silently lose their zeros — and validation systems downstream reject them. Google Sheets does the same thing on import unless you turn off automatic conversion — see our Google Sheets leading zeros guide. - Encodings get mixed up. A file saved as UTF-8 but opened as Windows-1252 turns every accented character into two garbage ones:
cafébecomescafé. Product names, customer names and addresses fill up withÃ,â€andÂsequences.
CSVUndo repairs all of these damage patterns after the fact — and each dedicated guide (scientific notation, lost leading zeros, mojibake) shows you the import technique that prevents the damage next time — see how to open a CSV in Excel without breaking it.
Who runs into this daily? E-commerce sellers uploading product feeds, accountants importing bank exports, logistics teams handling tracking numbers, HR staff with employee IDs, and anyone whose CSV passes through Excel on its way somewhere else.
Frequently asked questions
Is CSVUndo really free?
Yes — every tool on the site is free with no signup, no watermark and no row limit that matters for normal business files. There are no ads and no trackers on the site today. Weighing your options first? Compare us to paid converters on the CSVFix alternative page.
Is my data safe? Where does the file go?
Nowhere. The repair engine is JavaScript running in your browser tab. The file is never transmitted — you can even disconnect from the internet after the page loads and the tool keeps working.
Can you recover digits Excel already destroyed?
Honest answer: partially. Excel keeps 15 significant digits — if your value was longer and Excel saved the file, the digits past the 15th became zeros permanently. We expand the notation so the format is valid again (which fixes most upload rejections), and our guides show how to re-import from the original source correctly so it never happens again.
What file types are supported?
.csv, .tsv and .txt with comma, semicolon, tab or pipe delimiters — the delimiter is detected automatically, and quoted fields (values containing commas) are handled correctly.
Does it work with big files?
Files up to about 250 MB work in Chrome or Edge on a capable machine. Safari, Firefox, phones, and lower-memory devices get a lower ceiling, because a large file still needs several times its size in working memory, so the page lowers its own ceiling rather than crashing your tab. If a file is refused, the message names the limit that applied. Processing happens off the main thread, so the page stays responsive while a file is being repaired and you can Cancel at any point. The page warns you before it hits that ceiling, because the work runs on your own machine rather than a server; for anything larger than your browser’s ceiling, split the file first. Because nothing is uploaded, speed depends only on your own computer.