Does this send my query anywhere?

No. Tokenizing and reflowing the SQL runs entirely in your browser.

Does it understand every SQL dialect?

It targets the common subset shared by MySQL, PostgreSQL, and SQL Server: standard clauses, string literals, double-quoted or backtick-quoted identifiers, and line/block comments. Dialect-specific syntax (like PostgreSQL's $$-quoted function bodies) isn't specially handled.

Is this a real SQL parser?

No. It's a tokenize-then-reflow formatter: it splits the query into tokens (respecting quoted strings and identifiers), then reassembles them with clause-based line breaks and indentation. It has no notion of statement structure beyond recognizing clause keywords, so it can't validate that a query is semantically correct.

Why is content inside parentheses left on one line?

Subquery bodies, function arguments, and column-definition lists inside parentheses aren't recursively reformatted, since that would require real parsing of nested statement structure. Only top-level (paren-depth zero) clause keywords and commas trigger line breaks.

Can I keep my original keyword casing?

Yes. Toggle "Preserve case" to keep keywords exactly as typed instead of uppercasing them.

SQL Formatter - Format SQL Online

Overview

Introduction

A SQL query written on one line, or pasted from a compact source, is hard to scan for its clauses and conditions at a glance.

This tool tokenizes your SQL and reflows it so SELECT, FROM, WHERE, JOIN, GROUP BY, ORDER BY, HAVING, and LIMIT each start their own line, entirely in your browser.

What Is SQL Formatter?

A hand-written tokenize-then-reflow formatter: a scanner splits the query into keywords, identifiers, string literals, numbers, and punctuation (respecting quote boundaries and comments), then a reflow pass rebuilds the query with clause keywords on their own lines and indented continuation lines underneath.

It recognizes a broad set of standard SQL keywords, common data types, and common aggregate/scalar functions, uppercasing (or preserving) only those tokens; everything else, including string literals and identifiers, passes through untouched.

How SQL Formatter Works

The scanner walks the input character by character, treating single-quoted strings, double- or backtick-quoted identifiers, and `--`/`/* */` comments as opaque spans so nothing inside them is ever recased or split.

Adjacent keyword tokens matching known phrases (like GROUP BY, LEFT OUTER JOIN, or IS NOT NULL) merge into a single compound keyword before formatting, so multi-word clauses stay together.

During reflow, each top-level clause or JOIN keyword starts a new line at the base indent; its contents start an indented continuation line, further broken at top-level commas (in SELECT/GROUP BY/ORDER BY/SET) or AND/OR, while anything inside parentheses stays on one line.

When To Use SQL Formatter

Use it when a query pasted from a log, ORM debug output, or a single-line migration file needs to be readable during review.

It's also handy for normalizing keyword casing across a codebase that mixes SELECT and select inconsistently.

Often used alongside SQL to NoSQL Query Converter, JavaScript Formatter and Code Minifier.

Features

Advantages

Runs entirely client-side, so queries never leave your browser.
Recognizes a broad set of standard keywords, data types, and common functions rather than a handful of clauses.
Never touches string literals or quoted identifiers, so values like `'select this'` keep their exact casing.
Configurable indent width (2 or 4 spaces, or a tab) and an optional keyword-casing toggle.

Limitations

This is a tokenizer and line-reflow pass, not a real SQL parser: it doesn't validate query correctness, and it has no notion of nested statement structure beyond top-level clause keywords.
Content inside parentheses (subqueries, function call arguments, column-definition lists) is left on one line rather than recursively reformatted.
DDL statements (CREATE TABLE and similar) get keyword casing but not the same clause-per-line treatment as SELECT/INSERT/UPDATE/DELETE, since their structure lives inside a single parenthesized list.

Examples

SELECT with JOIN, WHERE, and GROUP BY

Input

select u.id, u.name from users u left join orders o on o.user_id = u.id where o.total > 100 and u.active = 1 group by u.id order by u.name

Output

SELECT
  u.id,
  u.name
FROM
  users u
LEFT JOIN
  orders o ON o.user_id = u.id
WHERE
  o.total > 100
  AND u.active = 1
GROUP BY
  u.id
ORDER BY
  u.name

Each clause starts its own line, the JOIN condition and WHERE/AND conditions indent underneath it, and SELECT's column list breaks at each top-level comma.

String literal containing a keyword-looking word

Input

select 'select this' as label from t

Output

SELECT
  'select this' AS label
FROM
  t

The word "select" inside the quoted string is left exactly as typed since it's inside a string literal, only the real SELECT keyword and AS are uppercased.

Best Practices & Notes

Best Practices

Format a query before pasting it into a code review or documentation so clauses are easy to scan.
Use the preserve-case toggle if your team's style guide lowercases keywords instead of uppercasing them.
Re-run a query through this tool after editing it by hand to keep clause indentation consistent.

Developer Notes

Tokenization is a hand-written character scanner (no external SQL parser library) that tracks quote and comment boundaries; a second pass merges adjacent keyword tokens matching known multi-word phrases (longest match first), and a third pass reflows the token stream into indented lines based on a fixed set of clause- and JOIN-keyword line breaks plus top-level comma/AND/OR breaks.

SQL Formatter Use Cases

Cleaning up a single-line query copied from a database log or ORM debug output
Normalizing SELECT/select keyword casing across a team before committing a migration
Making a deeply nested WHERE clause with several AND/OR conditions easier to read during a code review

Common Mistakes

Expecting subquery or function-argument content inside parentheses to be recursively reformatted; it's left on one line since this is a reflow pass, not a full parser.
Assuming the tool validates query correctness; it only re-formats tokens it recognizes and won't catch a missing FROM clause or a typo in a keyword.

Tips

If a query still looks cramped after formatting, it's likely because most of its content sits inside parentheses (a subquery or large IN list), which this tool intentionally leaves on one line.
Switch to 4-space or tab indentation if your team's SQL style guide doesn't use 2 spaces.

SQL Formatter