Write SQL with Gemini assistance Stay organized with collections Save and categorize content based on your preferences.

Preview — Gemini Code Assist

This feature is subject to the "Pre-GA Offerings Terms" in the General Service Terms section of the Service Specific Terms. You can process personal data for this feature as outlined in the Cloud Data Processing Addendum, subject to the obligations and restrictions described in the agreement under which you access Google Cloud. Pre-GA features are available "as is" and might have limited support. For more information, see the launch stage descriptions.

This document describes how you can use Gemini, an AI-powered collaborator in Google Cloud, to help you do the following in Cloud SQL Studio:

• Generate SQL queries using natural language prompts
• Explain SQL queries in the query editor

Learn how and when Gemini for Google Cloud uses your data.

This document is intended for database administrators and data engineers who are familiar with SQL, data analysis, and Cloud SQL. If you're new to Cloud SQL, see Cloud SQL overview.

Before you begin

1. To complete the tasks in this document, ensure that you have the necessary Identity and Access Management (IAM) permissions.
2. Confirm access to Cloud SQL Studio. If you don't have access to Cloud SQL Studio, then see Cloud SQL Studio required roles and permissions

3. Optional: To follow along with the examples in this document, create a Singers table using the following schema:

   CREATE TABLE
   Singers (
     SingerId   BIGINT PRIMARY KEY,
     FirstName  VARCHAR(1024),
     LastName   VARCHAR(1024),
     SingerInfo VARBINARY(MAX),
     BirthDate  DATETIMEOFFSET
   );
   

   After you create the Singers table, click refreshRefresh to update your database schema.

Required roles and permissions

To get the permissions that you need to complete the tasks in this document, ask your administrator to grant you the Cloud SQL Admin (roles/cloudsql.admin) IAM role on the project. For more information about granting roles, see Manage access to projects, folders, and organizations.

You might also be able to get the required permissions through custom roles or other predefined roles.

You also need the following permissions:

• Either serviceusage.services.list or serviceusage.services.get permissions for read capabilities
• cloudaicompanion.companions.generateChat permission to use explain SQL statements in the query editor
• cloudaicompanion.companions.generateCode permission to use help me code

Generate SQL queries using natural language prompts

As an early-stage technology, Gemini for Google Cloud products can generate output that seems plausible but is factually incorrect. We recommend that you validate all output from Gemini for Google Cloud products before you use it. For more information, see Gemini for Google Cloud and responsible AI.

You can give Gemini natural language comments (or prompts) to generate queries that are based on your schema. For example, you can prompt Gemini to generate SQL in response to the following prompts:

• "Add a date column called birthday to the Singers table."
• "How many singers were born in the 90s?"

If your schema changes, such as addition of a new table or column, then click refreshRefresh to update your schema before using Gemini.

To generate SQL in Cloud SQL with Gemini assistance, follow these steps:

1. In the Google Cloud console, go to the Cloud SQL page.

   Go to Cloud SQL

2. Select an instance from the list.

3. In the navigation menu, click Cloud SQL Studio.

4. Sign in to Cloud SQL Studio using the name of your database, username, and password. The Explorer pane displays a list of objects in your database.

5. In the taskbar, click settings_suggestionarrow_drop_downGemini to view Gemini features in Cloud SQL.

6. Ensure that Comment-to-query generation is enabled.

7. To query your database, click add New SQL editor tab.

8. To generate SQL, type a comment in the query editor starting with -- followed by a single-line comment, and then press Return or Enter.

   For example, enter the following prompt:

   -- add a row to table singers
   

   Press Return or Enter. Gemini generates SQL that's similar to the following:

   INSERT INTO Singers (SingerId, FirstName, LastName, BirthDate)
   VALUES (1, 'Alex', 'M.', '1977-10-16');
   

   To continue the example using the Singers table, enter the following prompt:

   -- show all singers born in the 70s
   

   Gemini generates SQL that's similar to the following:

   SELECT *
   FROM Singers
   WHERE Singers.BirthDate
   BETWEEN '1970-01-01' AND '1979-12-31'
   
   

   Note: Gemini might suggest different queries each time that you enter the same prompt.

9. Review the SQL suggestion and take any of the following actions:

   Caution: Because Data Manipulation Language (DML) and Data Definition Language (DDL) queries generated by Gemini can overwrite data, we recommend that you validate the generated queries before you run the queries.
   • To view the options for accepting the SQL generated by Gemini, hold the pointer over the query. The following options are displayed:
     • Accept: To accept the suggested query, press Tab, then click Run.
     • Accept word: To accept the suggested query partially, press Control+Right arrow. (or Command+Right arrow on macOS), then click Run.
   • To edit the original SQL, press Tab, edit the SQL, and then click Run.
   • To dismiss the suggestion, press Esc or continue typing.

Help me code tool

To use the Help me code tool, follow these steps:

1. In the Google Cloud console, go to the Cloud SQL page.

   Go to Cloud SQL

2. Select an instance from the list.

3. In the navigation menu, click Cloud SQL Studio.

4. Sign in to Cloud SQL Studio using the name of your database, username, and password. The Explorer pane displays a list of objects in your database.

5. To query your database, click the add New tab.

6. Click settings_suggestion Help me code next to the query editor.

7. In the Help me code window, enter a prompt. For example, add a row to table singers and click Generate.

   Gemini generates SQL that's similar to the following:

   INSERT INTO Singers (SingerId, FirstName, LastName, BirthDate)
   VALUES (1, Alex, 'M.', '1977-10-16');
   

   Note: Gemini might suggest different syntax each time that you enter the same prompt.

8. Review the generated SQL and take any of the following actions:

   Caution: Because Data Manipulation Language (DML) and Data Definition Language (DDL) queries generated by Gemini can overwrite data, we recommend that you validate the generated queries before you run the queries.
   • To accept SQL generated by Gemini, click Insert to insert the statement into the query editor. Then click Run to execute the suggested SQL.
   • To ask Gemini to generate a new query, click Edit. After you've edited your prompt, click Update. You can then decide to accept the new generated statement or dismiss the suggestion.
   • To dismiss the suggestion, close the Help me code window.

Explain SQL statements in the query editor

You can use Gemini in Cloud SQL to explain SQL queries in natural language. This explanation can help you understand the syntax, underlying schema, and business context for complex or long queries.

1. In the Google Cloud console, go to the Cloud SQL page.

   Go to Cloud SQL

2. Select an instance from the list.

3. In the navigation menu, click Cloud SQL Studio.

4. Sign in to Cloud SQL Studio using the name of your database, username, and password. The Explorer pane displays a list of objects in your database.

5. To query your database, click the add New tab.

6. In the query editor, paste the query you want explained.

7. Highlight the query that you want Gemini to explain, and then click astrophotography_mode Explain this selected query.

   The SQL explanation appears in the Gemini Cloud Assist pane.

Use Gemini to fix errors in queries

Preview — Gemini Code Assist

This feature is subject to the "Pre-GA Offerings Terms" in the General Service Terms section of the Service Specific Terms. You can process personal data for this feature as outlined in the Cloud Data Processing Addendum, subject to the obligations and restrictions described in the agreement under which you access Google Cloud. Pre-GA features are available "as is" and might have limited support. For more information, see the launch stage descriptions.

Gemini now offers an extended capability that helps you fix errors in your Cloud SQL queries without having to leave Cloud SQL Studio.

Gemini's AI-powered capabilities are integrated within Cloud SQL Studio to provide the following benefits:

• Provides a line-for-line comparison of the original query next to a new query with recommended changes using the Cloud SQL Studio editor.
• Highlights segments of the original query that might be causing the issue and provides a natural-language summary of changes.
• References the exact error message and schema information associated with the query to provide accurate, relevant corrections.
• Eliminates the need to rely on third-party, general-purpose LLMs that might provide less-nuanced help and might also expose your data to potential security risks.
• Provides faster, more efficient troubleshooting for common issues such as syntax, schema, and runtime errors.

Required permissions

Before you begin, confirm you have the following permissions:

• The permissions outlined in Required permissions.
• The Gemini for Google Cloud API permission, cloudaicompanion.googleapis.com/instances.generateCode.

Fix your query using Gemini and Cloud SQL Studio

To use Gemini's capabilities to fix certain errors in a query in Cloud SQL Studio, complete the following:

1. In the Google Cloud console, go to the Cloud SQL page.

   Go to Cloud SQL

2. Select an instance from the list.

3. In the navigation menu, click Cloud SQL Studio.

4. Sign in to Cloud SQL Studio using the name of your database, username, and password. The Explorer pane displays a list of objects in your database.

5. In the taskbar, click settings_suggestionarrow_drop_downGemini to view Gemini features in Cloud SQL.

6. To query your database, click the add New tab.

7. In the query editor, enter a query that might be incorrect or incomplete.

8. Click Run.

   If the query contains an issue, an error is returned.

   For example, the message returned might be a validation error in the Run tab, or an error message in the Results tab.

9. Click Fix.

10. In the query window, a differences editor ("diff editor") opens in the tab, showing a line-for-line comparison of the existing query with recommended changes. The query editor also provides a summary of the suggested edits. Review the suggested changes to the query and update as needed. Then, accept and run your new query:

    Click Accept and run.

Limitations

• This capability is only accessible within Cloud SQL Studio in Google Cloud console.
• Current Gemini context window limits apply.

What's next

• Read Gemini for Google Cloud overview.
• Learn how Gemini uses your data.
• Learn how to analyze your data in Google Cloud using Gemini.