Skip to main content
All checkers

CSV Validator

Check a CSV on your device and find the records to fix before import.

Your CSV

Up to 5 MB. Nothing is uploaded.

Format

First record

Column rules

Enter exact column names, one per line.

One per line. Use column | type, then add | required for a value check.

Leave blank to skip the duplicate check.

Your CSV check

Example inputs
Validation verdict
Checks failed

3 data records checked. 1 field issue and 1 duplicate group.

  1. Record 3, line 3, column 2. Enter a valid decimal.
  2. Duplicate group 1. Record 2, line 2, column 1.
  3. Duplicate group 1. Record 4, line 4, column 1.

Choose the CSV Rules to Check

A CSV validator checks how the file breaks into records and columns. It then checks the values against the column rules you enter. Use the requirements for your own import when you select those rules.

Set the delimiter to comma, semicolon, tab or pipe to match your source. Choose UTF-8, UTF-16 LE or Windows-1252 when reading a file. Pasted text is already decoded.

Choose Headers if the first record contains column names. Choose Data to start checking values immediately and refer to columns as Column 1, Column 2 and so on.

In the example, 001,1200,"Branch, west" has three fields. Quotes keep the comma inside the third field, and two quotes inside a quoted field stand for one literal quote. A quoted field can span lines under the RFC 4180 quoting rules.

Enter the required column names on separate lines. For a value check, enter Amount | decimal | required under Field rules. The last part flags empty or whitespace-only values and is optional.

Use text for any populated text, integer for whole numbers, decimal for numbers with an optional decimal point or date for calendar dates in YYYY-MM-DD form. Number rules allow a plus or minus sign and omit thousands separators and currency symbols.

Enter an identifier column to find repeated IDs. The exact match treats 001 and 1 as different values and flags empty or whitespace-only identifiers. These are your import checks, and each lender or system can set its own requirements.

The validator accepts up to 5 MB per file, five million text characters, 10,000 data records and 100 columns.

Read Row and Field Diagnostics

Start with the reported record and column. A quoting error must be fixed before later record boundaries can be checked. An inconsistent field count points to the first extra or missing column, and missing required columns are named separately.

Record 1 is the header when Headers is selected and the first data record otherwise. A quoted multiline field stays within one record. The line number marks the physical line where the record starts.

In the example, record 3 has a value that fails the decimal check in column 2. Records 2 and 4 share duplicate group 1 in column 1. The report groups their locations without including the identifier or other values from those records.

Correct the Source and Validate Again

  1. Save a copy in the application where you maintain the data.
  2. Correct the reported fields and export the corrected copy with the intended delimiter and encoding.
  3. Choose the corrected file and run the same format and column rules again.

A clear report means the CSV passes the selected checks. Run the destination system’s validation before importing to catch its business rules and confirm that it accepts the file.

Keep Sensitive Records on Your Device

The browser reads the file or pasted text in this tab. The CSV and record-level diagnostics stay on your device. Clear CSV removes the selected file, pasted text, column rules and report from the tool and restores the default format settings.

Download report saves a JSON file with the verdict, issue descriptions and locations. It keeps duplicate records together by group number. The download stays on your device after you clear the tool.

The validator leaves your source unchanged, including date formats, identifier spaces and leading zeros. Its only export is the diagnostic report.