CSV editor — fix values without Excel
Open a CSV in an editable grid, correct the cells that need fixing, and download the cleaned file. No Excel, no upload, no reformatting surprises.
Click to upload or drag and drop
Drop your CSV, Excel, or JSON file
Max 10MB
Runs 100% in your browser — this page makes no network request with your data.
- 1
Drop your CSV — it opens as an editable grid (first 500 rows editable).
- 2
Click any cell to edit; values stay exactly as typed (no date/zero mangling).
- 3
Download the edited CSV, or send it straight to analysis.
Worked example: fixing product SKUs before a catalog import
An ecommerce catalog manager receives a supplier price list as a CSV where a handful of SKUs are wrong and several product codes have leading zeros. Opening it in Excel is the usual trap: it strips the zeros and turns long codes into scientific notation the moment the file opens.
Instead they load the CSV into the editor, which shows it as an editable grid and treats every value as text exactly as typed. They click into the wrong cells, correct the SKUs, fix a couple of misspelled product names, and leave the zero-padded codes untouched and intact. Then they download the cleaned file.
Because the first 500 rows are editable in the grid, this quick-fix workflow suits targeted corrections rather than bulk transformations. For a rules-based change across every row, say prefixing all SKUs, they can send the file to the workspace instead, then turn the cleaned data into a report.
A CSV editor lets you open a comma-separated file, change the cells that need fixing, and save it back, ideally without altering anything you did not touch. To edit a CSV online, open it in a browser grid like the one above, click the cells to correct, and download the result; the values stay exactly as typed because the editor treats every cell as text.
That last point is the whole reason to edit a CSV somewhere other than Excel. A spreadsheet program rewrites values it thinks are numbers or dates the instant the file opens, so a quick fix can introduce ten new problems you never see until an import fails downstream.
What Excel silently changes when you open and save a CSV
Excel does not wait for you to edit anything. It guesses the type of every column on open and reformats accordingly, and when you save, those guesses overwrite the original text permanently. You can open a clean file, change nothing, save it, and hand back a corrupted export.
These are the changes it makes without asking:
- Leading zeros are stripped: 01730 becomes 1730, breaking ZIP codes and zero-padded IDs.
- Long numbers become scientific notation: a 13-digit barcode or a long order ID collapses to something like 1.05E+11, losing the exact digits.
- Date-like text is converted: values that resemble dates are rewritten into Excel's date serials.
- Trailing decimal zeros disappear: 1.50 becomes 1.5, which matters for price and version fields.
- Encoding is re-guessed on save, sometimes turning a clean UTF-8 file into one with garbled accented characters.
When a browser CSV editor beats Excel
Not always, and it is worth being honest about that. Excel wins for formulas, pivot tables, and heavy reshaping. A browser editor wins for the specific job of making small, faithful corrections to a file that has to import somewhere else exactly as written.
Reach for an online editor when text fidelity matters more than spreadsheet power:
- Fixing a handful of wrong values, a typo in a name, a bad email, a mislabeled category, without risking the columns you are not touching.
- Preserving IDs, ZIP codes, phone numbers, and SKUs that Excel would mangle.
- Working on a locked-down laptop or Chromebook with no Office license and no install rights.
- Editing a file that came in as UTF-8 and must go back out as UTF-8 unchanged.
The scope of a browser grid
The editor above opens the file as an editable grid and treats every value as the text you typed, so nothing is reformatted on open or download. The first 500 rows are editable in the grid, a deliberate performance boundary rather than a hidden limit; it suits targeted corrections. For a rules-based change across every row, the next section covers your options. When you are done, spot-check the result in the CSV viewer, break it into import-sized parts with the CSV splitter if it is too large to upload in one pass, or send it straight to analysis. If you are correcting several files to combine, fix each and then join them with the CSV merger.
Bulk find and replace across a CSV
When the fix is the same across many rows, standardizing a country code, replacing a delimiter inside a field, cleaning a recurring typo, cell-by-cell editing is the wrong tool. You have three practical routes, in rising order of power.
A plain text editor like VS Code or Notepad++ does find-and-replace with regular expressions and shows you the raw file, which is perfect for a literal string swap. The caution is that a CSV is structured: a naive replace can hit characters inside quoted fields you meant to leave alone, so anchor your pattern carefully and check the diff. The command line sed does the same at scale for the scriptable. And when the change is genuinely rules-based, say, prefix every SKU, or normalize every date to ISO format, describe it in the workspace and let real Python apply it across all rows at once, which also handles quoting and encoding correctly.
Preserving formats for re-import into CRMs and ESPs
Most CSV edits exist to feed another system, a CRM, an email service provider, an ad platform, and those systems are strict about format. An edit that reads fine to you can still be rejected on import if it violates the destination's expectations. Get these right before you upload.
| Field type | Keep it as | What breaks it |
|---|---|---|
| Dates | One consistent format, ideally ISO 8601 (2026-07-18) | Mixed or locale-flipped formats (03/04 read as the wrong month) |
| Phone numbers | Text, often E.164 like +14155550100 | Excel stripping the plus or a leading zero |
| IDs and ZIP codes | Text with leading zeros intact | Auto-conversion to numbers or scientific notation |
| Booleans | The exact tokens the target expects (yes/no, TRUE/FALSE) | Silent conversion to 1/0 or checkboxes |
| Encoding | UTF-8 | Accented characters garbled by a Windows-1252 save |
Frequently Asked Questions
Everything you need to know about using AnalyzeData.
Excel silently reformats data on open — stripping leading zeros, converting IDs to scientific notation, rewriting dates. This editor treats every value as text you typed, so the file round-trips faithfully.
Files of any size open; the first 500 rows are editable in the grid (a browser-performance boundary we state rather than hide). For bulk transformations, the analysis workspace is the better tool.
No — editing happens in memory in your browser, and the download is generated locally.
Edit the file here rather than in Excel. This editor treats every cell as the text you typed, so zip codes like 01730, zero-padded SKUs, and long numeric IDs stay exactly as they are instead of being stripped or converted to scientific notation. When you download, the file round-trips faithfully; the values you saw are the values you get.
Yes. The grid opens in your browser and applies no automatic formatting, so dates, currencies, and IDs are left as-typed and nothing gets silently rewritten on open or save. It works the same on Windows, Mac, or a Chromebook without installing anything, and the file is edited in memory and never uploaded while you work.
Yes, if you use an editor that treats values as text rather than re-typing them for you. A browser CSV editor changes only the cells you click and leaves every other value byte-for-byte identical on download. Excel is the opposite: it can reformat columns you never touched the moment the file opens, so a single edit there risks silent changes across the whole file.
Open the file in an online CSV editor, click the cell, correct it, and download. Because the grid applies no automatic formatting, the fix is the only change, and dates, IDs, and ZIP codes elsewhere stay exactly as they were. It works on any device with a browser, with no Office license and nothing to install.
Edit it as text, preserve the destination's expected formats for dates, IDs, and booleans, keep the encoding as UTF-8, and run one round-trip test on a copy before processing the full file. A browser editor helps because it will not silently reformat values, so what you see on download is exactly what the target system receives.
Not efficiently by hand. For a repeated change, use find-and-replace in a text editor for a literal swap, or describe the rule in an analysis workspace and let code apply it to every row, which also keeps quoting and encoding correct. Cell-by-cell editing is for targeted fixes; rules-based changes belong in a tool built to transform the whole file.
When a table isn't enough
The workspace runs verified AI analysis on the same file and turns the results into a report you can send.
Analyze this data instead