LumaSoft products — AI tools for database professionals

We build AI tools that run right inside your database client. Meet the SqlSage family.

  • SqlSage for DBeaver

    For teams on DBeaver across many databases — PostgreSQL, MySQL, Oracle and more, on any OS.

    AI for queries and data work inside DBeaver — without leaving your editor.

    Learn moreOpen product

  • SqlSage for SSMS

    For SQL Server pros working in SSMS on Windows.

    AI for T-SQL and everyday SQL Server work — right inside SSMS.

    Learn moreOpen product

Frequently asked questions about SqlSage

What is SqlSage?

SqlSage is an AI assistant for working with SQL that runs right inside the tool where you already write queries — SQL Server Management Studio (SSMS) and DBeaver. Instead of switching to a separate chat window, you talk to the assistant inside your editor. It grounds its answers in the real schema, values and query plans from your database, true to the “prove, don’t guess” principle.

Does SqlSage work inside SSMS and DBeaver?

Yes. SqlSage is a plugin for your database client, not one more separate window: an add-in for SQL Server Management Studio and a plugin for DBeaver. The SSMS edition targets Microsoft SQL Server on Windows; the DBeaver edition works across many database engines from a single client.

Does SqlSage require an API key?

No. SqlSage runs keyless — you connect your own Claude, ChatGPT or Codex account. There is no API key to buy and no second AI subscription; you pay only for the model you already use.

Does SqlSage send my data to the AI model?

It does not send the contents of your tables. The model receives the database schema and the SQL, not the data in your tables. On top of that, SqlSage is read-only by default: statements that change data, and DDL, run only after you explicitly confirm them. In DBeaver, where the engine allows it, read-only is also enforced on the server side.

Which database engines does SqlSage support in DBeaver?

SqlSage for DBeaver has dedicated, dialect-aware support for the five most common engines: PostgreSQL, MySQL/MariaDB, Oracle, SQL Server and SQLite. It knows their paging syntax, identifier quoting and how each one exposes a query plan, so its suggestions match the database you are actually on.