Jam SQL Studio

Jam SQL Studio is a free, cross-platform SQL IDE for SQL Server, PostgreSQL, MySQL, MariaDB, Oracle and SQLite, built as an Azure Data Studio replacement. It pairs a full query editor with execution plans, schema and data compare, notebooks and a local MCP server that lets AI agents inspect and query databases under explicit approval.

Jam SQL Studio is a desktop SQL IDE that connects to SQL Server, Azure SQL, PostgreSQL, MySQL, MariaDB, Oracle and SQLite through one editor. It is positioned as the successor to Azure Data Studio, which Microsoft retired in February 2026, and it runs on macOS, Windows and Linux with a consistent interface across all three. The pitch is straightforward: teams that lost their long-standing client get the same connection files, .sql scripts and .ipynb notebooks back without relearning a workflow. The practical work happens in the editor. Smart IntelliSense adapts to the connected engine, a Table Explorer allows inline editing with foreign key lookups, and per tab transaction control lets a developer roll back a single window without touching the rest of the session. Beyond query writing, the tool covers visual execution plans (tree and graph), schema compare with generated ALTER and CREATE sync scripts, data compare with row level diffs, built-in charts, and a Database Blueprint that stores schema as a Git versioned folder of .sql files. DBA functions such as session browsing, a performance dashboard, backup and restore, and filegroup or tablespace management are present for four engines. The AI layer is the differentiator, and it is deliberately conservative. A local MCP server on 127.0.0.1 exposes connection and UI control tools to external agents under bearer token auth, with no queries run by default and writes gated behind per statement approval. An AI Workspace generates CLAUDE.md and AGENTS.md context files plus a schema export so agents understand the database before writing anything, and an in-app sidebar connects to a locally installed Claude Code CLI using the user's own subscription. All AI features ship disabled. It suits developers, analysts and DBAs who work across several engines or who want agent tooling without handing a remote service their credentials. Compared with DBeaver, DataGrip, Navicat and TablePlus, the edge is the MCP integration and the free personal tier; the tradeoff is a younger product with a shorter track record than the incumbents.