Generate SQL
Give SQL Mocker the database context it needs by extracting or uploading schema metadata. Review the structure, ask a question in natural language, and generate SQL without connecting AI directly to your database.
Select your database system and the tables you want to use. SQL Mocker generates a metadata extraction script that you run inside your own database tool. Paste or upload the results to bring in table names, columns, data types, keys, and recorded relationships without sharing database credentials.
If you already have schema metadata, upload the file directly. SQL Mocker turns it into a reviewable workspace where you can check tables, columns, keys, and table connections before using that structure to generate SQL.
How it works
Four guided steps, with no direct database connection.
Generate a metadata script to run in your own database, or upload an existing schema or metadata file.
Check tables, columns, keys, and relationships. Confirm or add connections where metadata is incomplete.
Ask a question in natural language and use schema-field suggestions to select the correct tables and columns.
Inspect the generated query, preview results with dummy data, and copy the SQL to run in your own database.
Prompt assistance
Click in the prompt box and describe the data you need. If you are unsure of the exact table or column names, schema-field suggestions search the available database structure as you type.
Add your database structure, review the relationships, and ask your first question.
Use the demo dataset