CSV Tools
CSV Validator
Check a CSV table before importing it. Find broken quoted fields, repeated or empty headers, and records with the wrong number of columns.
✓ Every tool runs in your browser. We do not upload your input or output.
Keyboard shortcuts
Ctrl/⌘ + Enter: process
Ctrl/⌘ + Shift + F: full screen
Esc: exit full screen
Ctrl/⌘ + F in editor: search
How to use
- Paste CSV text, upload a file, or load the sample table.
- Choose the delimiter used by the source: comma, semicolon, or tab.
- Read the validation result or select an error to jump to its line and column.
- Correct the source and validate again before using the file in another system.
Examples
Check a consistent table
The header names are unique and both records contain two fields.
id,name A-101,Avery A-102,Morgan
Valid CSV: 2 data rows, 2 columns.
Catch a missing cell
The final record has fewer fields than the header and is reported at its starting line.
id,name A-101,Avery A-102
Row 3 has 1 fields; expected 2. (line 3, column 1)
Find structural problems early
A CSV file can look aligned in a text editor while still having a broken record. A comma inside an unquoted note adds an extra field. An unclosed quotation mark can absorb the next line into one cell. A duplicate header can cause one value to overwrite another after conversion. The CSV validator checks these structural rules before you pass the table to a spreadsheet, script, or import pipeline.
Start by choosing the delimiter that the source file actually uses. The tool supports commas, semicolons, and tabs. It treats the first record as a header and checks that every name is nonempty and unique. Each following record must have exactly the same number of fields. Quoted delimiters, doubled quotes, and line breaks inside quoted fields are parsed as cell content rather than separators.
Use the location as a guide
When a quote or header is malformed, the error includes the physical line and column where the parser found the problem. For a short or wide record, the reported line is where that record begins. Select the error to move the editor cursor to the location. A malformed quote may affect later lines, so also inspect the previous field when the marked row appears correct.
Validation does not infer data types or application rules. 0012 and 12 are both valid CSV text, even if a receiving system treats them differently. Empty data cells are allowed; missing separators are not. Once the table passes this structural check, use the destination’s schema or import preview to confirm the meaning of each column. All parsing happens locally, and the workspace panels scroll internally for large files.
Frequently asked questions
Does valid CSV mean the data is correct?
No. This checks table structure, not whether names, dates, or IDs meet your application's rules.
How are line numbers counted for multiline cells?
Errors use physical input lines. A quoted cell can span lines without creating extra records.
What happens with duplicate headers?
The second occurrence is reported because converters cannot create unambiguous field names from it.
Are blank lines accepted?
Empty physical lines outside quoted fields are skipped. An empty quoted cell remains a real cell.
Does the validator upload my CSV?
No. Validation runs in a Web Worker inside your browser.