Convert CSV to PostgreSQL

Upload a .csv or .tsv file and get back a PostgreSQL SQL script you can load straight into your database. Types are preserved rather than flattened to text. Free, no registration, up to 10MB.

Click to upload or drag and drop

CSV, Excel, SQLite, SQL, DBF (max 10 MB)

How your data is typed in PostgreSQL

Generated from the converter's own PostgreSQL type mapping — this is what it actually emits, not an example.

Your dataPostgreSQL type
Whole numberBIGINT
Decimal (money)NUMERIC(10,2)
Floating pointDOUBLE PRECISION
True / falseBOOLEAN
DateDATE
Date and timeTIMESTAMP
Short textTEXT
Long textTEXT
Binary dataBYTEA

Working from a .csv file

A CSV file carries no header describing itself, so everything about it - the separator, the encoding, whether an empty field means nothing or means zero - has to be worked out from the bytes. That is where the risk in this conversion lives.

  • →The separator is sniffed from the first 4 KB across comma, semicolon, tab and pipe, so a semicolon-separated file saved by Excel in a European locale needs no setting.
  • →The UTF-8 byte-order mark Excel writes is stripped before the header row is read, so your first column is not named with an invisible character in front of it.
  • →Types are decided from the first 10,000 data rows. A value further down that contradicts the type it settled on becomes NULL rather than failing the whole file.
  • →The table is named after the file, with spaces and hyphens replaced by underscores: sales report.csv becomes sales_report.

What the PostgreSQL script contains

The conversion produces one dump.sql with a README beside it, zipped. It is a script you can open, grep and edit before you run it rather than an archive a tool has to unpack, and it is deliberately sparing in what it declares.

  • →Every string column is declared TEXT, whatever length the source declared for it. PostgreSQL stores a short value in a TEXT column exactly as cheaply as in a VARCHAR, and an unbounded column cannot reject a row on load for being a few characters too long.
  • →Identifiers are all double-quoted, and even so a name on PostgreSQL's reserved list is given a trailing underscore: a column called user arrives as user_, order as order_, limit as limit_.
  • →A decimal column whose source declared no scale is given six decimal places rather than none, keeping whatever precision it did declare: NUMERIC(10,6) from a source that said 10, and NUMERIC(18,6) from one that declared neither. A scale of zero would have the server round 19.99 to 20 as it loaded.

Reading CSV

  • →Leading zeros are preserved. A column containing 01234 is kept as text, never turned into the number 1234.
  • →A row with more fields than the header, usually an unquoted comma inside a value such as Acme, Inc., is left out and reported by line number instead of sliding every later value one column right. A trailing comma on every line adds no column.
  • →An empty cell becomes NULL, not an empty string — the two mean different things once the data is in a database.
  • →Amounts in accounting style, 1,250.00 with a thousands separator or (84.30) for a negative, become exact decimals when the column also has a decimal point; a lone 1,250 stays text because in much of Europe it means 1.25.
  • →A file that is not UTF-8 is read as Windows-1252, unless the text reads as Russian in Windows-1251 or DOS 866, which are checked as well; the warning names the code page chosen.
  • →Dates are only treated as dates when the format is unambiguous. 03/04/2025 could be March 4th or April 3rd, so it stays text unless another value in the column settles it.

Writing PostgreSQL

  • →Backslashes are left alone, because PostgreSQL treats them as ordinary characters — doubling them would corrupt every Windows path.
  • →Auto-increment columns become GENERATED BY DEFAULT AS IDENTITY.
  • →Binary data is written as bytea, and identifiers respect the 63-character limit.
  • →All-uppercase names from dBase, FoxPro or Oracle, such as CUSTNO, are written in lower case, so queries need no double quotes. Mixed-case names such as OrderID keep their spelling.

What you get

A ZIP containing dump.sql, plus a README with the exact command to load it. If anything about your data needed a judgement call, a _warnings.txt explains it.

Load this dump into PostgreSQL:

  psql -U USER -d DATABASE -f dump.sql

Create the database first if it does not exist:

  createdb -U USER DATABASE

Questions

Is converting CSV to PostgreSQL free?
Yes. Files up to 10MB convert free with no account. There is a limit of five conversions per day per IP address.
Will my CSV data keep its types in PostgreSQL?
Yes — that is the point. Whole numbers, decimals, dates and booleans are mapped to real PostgreSQL types rather than everything becoming text. The mapping table above shows exactly what you get.
Do you store my file?
No. The file is written to a temporary directory, converted, and the directory is deleted as soon as your download has been sent.

This conversion, under other names

Convert CSV to something else

Convert something else to PostgreSQL