Understand SQL in plain English with AST-based analysis, logical execution order, and clause-by-clause breakdown.
Start with the Explainer to understand intent, use the Validator to confirm syntax, then run the Linter for quality and safety heuristics before executing against a database.
SELECT c.country, COUNT(o.id) AS orders FROM customers c INNER JOIN orders o ON c.id = o.customer_id WHERE o.status = 'paid' GROUP BY c.country HAVING COUNT(o.id) > 5 ORDER BY orders DESC LIMIT 10; -- Summary: Retrieves paid order counts per country, -- keeps groups with more than 5 orders, returns top 10.
An SQL Query Explainer converts SQL text into plain-English descriptions of what the query does, how clauses relate, and in what logical order the database processes them. It helps you understand unfamiliar or complex SQL without running it.
No. It never connects to a database and never runs your queries. All analysis is static text parsing in your browser.
No. Parsing and explanation run locally. Your SQL stays on your device.
PostgreSQL, MySQL, MariaDB, SQLite, SQL Server (T-SQL), BigQuery, Snowflake, Redshift, DB2, Hive, and generic SQL via AST parsing with dialect mapping. Complex dialect-specific syntax may parse partially.
Yes. SELECT with JOINs, WHERE, GROUP BY, HAVING, ORDER BY, LIMIT, CTEs, and window functions is the primary focus with full clause and execution-order breakdown.
Yes at a basic level: target tables, SET/VALUES clauses, and WHERE filters are explained. DML explanations focus on intent and row impact, not execution plans.
Partially. Procedural blocks (PL/SQL, T-SQL batches) may parse with limited detail. A banner indicates when explanation coverage is partial.
For SELECT queries, SQL is logically processed as WITH → FROM/JOIN → WHERE → GROUP BY → HAVING → SELECT → DISTINCT → ORDER BY → LIMIT. This differs from the order you write clauses in the query text.
No. Database EXPLAIN shows actual execution plans and costs. This tool shows logical clause order and human-readable descriptions only — no index usage or row estimates from a live database.
It includes educational heuristics (SELECT *, missing WHERE on DELETE, leading wildcard LIKE) reused from the SQL Linter. These are not execution-tested performance metrics.
No. Without a schema connection, the tool cannot verify that tables or columns exist. It only describes what the SQL text references.
Beginner (simple language), Intermediate (balanced default), and Technical (includes clause terminology prominently).
Yes. Statements split on semicolons outside quoted strings are explained individually with a statement selector in the results panel.
Heuristic scores from 0–100 based on JOIN count, CTE count, nesting depth, line count, and style signals. They estimate learning difficulty, not runtime cost.
Yes. Copy or download TXT, Markdown, or JSON reports from the results panel.
Yes. Hand off to the SQL Validator for syntax checks, SQL Linter for quality review, or SQL Formatter to beautify output.
Small and medium scripts explain instantly. Inputs above ~300 KB use a Web Worker. Very large dumps may slow the browser; multi-gigabyte files are not supported in-browser.
Yes. After the page loads, explaining works offline because processing is client-side.
No. It is heuristic. Confirm the dialect manually when accuracy matters for dialect-specific syntax.
Yes when the AST parser recognizes them. WITH clauses appear first in the execution timeline; window functions are noted in the SELECT phase.
Understanding intent first reduces mistakes with destructive statements, helps code review, and speeds onboarding when reading legacy or AI-generated SQL.
The Validator checks whether SQL parses correctly and reports syntax errors. The Explainer describes what valid (or partially valid) SQL is intended to do in plain English.
The Linter flags style and safety issues with scores. The Explainer teaches structure and logic. Use both for a complete pre-flight review.
Yes. Enable auto-explain for debounced live analysis, or click Explain SQL / press Ctrl+Enter for immediate results.
Ctrl+Enter on Windows/Linux or Cmd+Enter on Mac.
Modern Chromium, Firefox, Safari, and Edge with Web Worker and localStorage support.
Yes. No account, usage limit, or per-query charge.
Yes. Beginner depth mode and execution-order steps are designed for learners reviewing JOINs, GROUP BY, and subqueries.
No. Explanations are generated deterministically from the parsed AST using local templates — no external AI API calls.
You will see parse errors with a link to open the SQL Validator. Fix syntax first, then re-run Explain.
Yes. Explaining AI-generated SQL before execution is a common use case to verify intent and catch risky patterns.