RegSQL vs Sqruff

RegSQL is a web SQL workspace combining browser-local formatting with optional AI assistance. Formatting runs in a browser worker without an account or server round trip. It adjusts whitespace and casing while preserving query logic, aliases, and comments, and accepts multi-statement input up to 1 MB. Listed dialects are PostgreSQL, MySQL, MariaDB, SQLite, BigQuery, SQL Server, Snowflake, and Oracle PL/SQL. AI tools can generate SQL from a description, explain an existing query, or suggest a repair. Formatting may reveal parsing problems, but RegSQL does not connect to a database or verify tables, columns, permissions, data types, or runtime behavior. SQL, prompts, and schema text are sent to the AI provider only after Generate, Explain, or Repair is chosen. The Free plan is free, billed Forever, with unlimited local formatting and 20 AI requests monthly. Indie costs $8 a month or $72 billed yearly ($6/month).

Sqruff is a free SQL linter and formatter written in Rust, focused on formatting valid SQL for particular dialects and providing linting and fixes. It offers configurable automated formatting, custom rules, and column-level data lineage analysis for SQL queries. Dialects include ANSI, Athena, BigQuery, ClickHouse, Databricks, DuckDB, MySQL, Oracle, Postgres, Snowflake, SQLite, and Trino, among others. Dialect and rule settings are read from a project `.sqruff` configuration file. Developers can use the maintained VS Code and Zed integrations, or an LSP server with other LSP-capable editors. A documented GitHub Action supports use in CI, and a browser playground is available. Installation options include Homebrew, a bash installer, pip, cargo, binary releases, and an official Docker image. The public repository provides an Apache-2.0 license. The project identifies dbt project support as experimental and is associated with Quary, an open-source analytics platform for engineers based in the United Kingdom.