Code · Made a little simpler

Data Type Guesser

Inspect CSV and JSON columns for numbers, text, dates, booleans and nulls. Review mixed values, quoted fields and precision risks locally.

Free to useNo account neededProcessed on your device
Your workspaceProcessed on your device

Add this tool to your website

Paste this code into your page to add the tool and a visible ToolMellow credit link.

HTML embed code
Preview embed

Data Type Guesser: how to use it

  1. Paste a delimited table with headers or a JSON array of objects. Select the input format and the field separator if using CSV.
  2. Choose whether quoted CSV fields should count as text. Select Inspect types and compare each suggested type with the counts and examples.
  3. Check leading zeros, precision risks and blanks before deciding which fields to import as text. Copy or download column-types.json.

Limits and notes

Runs on your device and leaves the source values intact. Inspects up to 500,000 input characters, 10,000 rows and 128 columns. Supports commas, semicolons, tabs and pipes for delimited text. JSON rows must be objects and keys must be unique.

How your privacy works →

Data Type Guesser: common questions

Are CSV quotes a declaration of type?

CSV uses quotes to delimit a field. The workspace offers a conservative setting that counts quoted fields as text. Clear it to look for boolean, number, date and null patterns inside the quotes.

What does a number precision flag mean?

The numeral is outside the JavaScript safe integer range or would change when parsed and written back as a JavaScript number. The evidence keeps its source spelling.

Are a blank CSV cell, JSON null and an absent JSON key equivalent?

The report counts Blank, Null and Missing separately. They do not outweigh a type observed in nonempty cells.

Guides