Skip to content

CSV Schema Inspector

Data & CSV · Done in 3 steps

Stays on your deviceFree · No sign-up
  1. 1Paste your CSV
  2. 2
  3. 3
Loading tool…

About

Profile every CSV column: inferred type, nulls, unique values, min/max, lengths and samples, plus a suggested SQL CREATE TABLE.

How to use it

  1. 1Paste CSV or drop a file.
  2. 2Review each column’s inferred type, null count, unique count, range and samples.
  3. 3Pick a SQL dialect and table name.
  4. 4Copy the CREATE TABLE statement or download the profile as CSV.
You give
CSV/TSV with a header row
You get
Column profile table and SQL DDL (PostgreSQL, MySQL, SQLite, SQL Server)
Good to know
Type inference recognises integers, decimals, booleans (true/false/yes/no), ISO and M/D/YYYY dates and datetimes; anything else is text. Values like “NA”, “null” and “-” count as nulls. Files up to roughly 50–100 MB work on a desktop browser (big files are streamed in chunks); phones may run out of memory sooner. The preview shows the first 200 rows, downloads include every row.
Privacy
Processed on your device. Nothing is uploaded, stored or logged.
Last tested
2026-10-28

FAQ

How is VARCHAR length chosen?

From the longest value seen, rounded up to 32, 64, 128 or 255; longer columns become TEXT. Add headroom if the data will grow.

Why is a numeric-looking column typed as text?

Either some values aren’t numbers (the inspector shows the mix) or they have leading zeros like zip codes, which should be stored as text.

Go further with AI