With Google Workspace Studio, you can automate routine tasks across Google Workspace services like Gmail, Sheets, and Drive, as well as other common business apps using Gemini.
Every agent has 2 main parts:
- Starter: This is the event that starts your agent. A starter can be a schedule, like “every Friday at 5 PM,” or an event, like to receive a new email with an attachment. Each agent has only one starter.
- Steps: These are the tasks the agent performs after it starts. You can add multiple steps to an agent. A step could be:
- “Draft a reply in Gmail.”
- “Add a row in Google Sheets.”
- “Send a message in Google Chat.”
For example, you can build an agent to:
- Save attachments from your emails to a specific folder in Google Drive.
- Send you a notification in Google Chat whenever someone submits a Google Form.
- Analyze new emails and give you an AI-generated summary.
Get started with Google Workspace Studio
Understand Workspace Studio requirements
- Device requirement: You must use a computer with a supported web browser to create agents. Once active, agents can run on any device.
- Google Workspace edition: To ensure that your edition supports Workspace Studio, check with your admin. To use AI features, your Google Workspace admin must enable Gemini on your account.
- Google Workspace availability:
- Business Starter, Standard, and Plus
- Enterprise Standard and Plus
- Education Fundamentals, Standard, Plus, and the Teaching and Learning add-on
- Google AI availability:
- Google AI Pro for Education
- Google AI Ultra for Business
- Google Workspace availability:
- App access: The templates available to you rely on your Google app access. For example, if your account doesn't have Gmail enabled, templates that use Gmail steps won't appear.
- Workspace Studio integrates with:
- Gemini
- Gmail
- Google Calendar
- Google Chat
- Google Drive
- Google Docs
- Google Forms
- Google Sheets
- Google Tasks
- Third-party apps like:
- Asana
- Mailchimp
- Salesforce
- Workspace Studio integrates with:
- Age restriction:
If you're signed in with a school account and are under 18, you can’t use any AI features in Workspace Studio. If an agent containing AI steps is shared with you, those steps will be removed when you open the agent.
Tutorial: Create your first agent
The easiest way to learn how Workspace Studio works is to start with a template.
This example guides you to set up an agent that uses Gemini to review your unread emails and send you a daily summary in Google Chat.
Step 1: Choose a template
- On your computer, go to studio.workspace.google.com.
- On the "Discover" page, check the list of pre-built agent templates.
- Click Get a daily summary of unread emails template.
- This opens the agent editor that you can use to configure your agent.
Step 2: Configure your agent
Next, tell the agent exactly what to do. For this template, the logic is already built for you, but it's helpful to click through the steps to know how your agent works.
- Step 1: On a schedule. This first step is the starter and sets when the agent runs. By default, it may be set to run every weekday morning. You can change the time or frequency to fit your routine.
- Step 2: Recap unread emails. This is an AI-powered step. It's configured to look at your emails from "Yesterday" and use Gemini to generate a summary. You can leave the prompt and settings as they are.
- Step 3: Notify me in Chat. In the "Message" field, you’ll see a blue variable chip named [Step 2: Summary of unread email]. This tells the agent to take the AI summary created in the previous step and send it to you as a chat message.
Step 3: Test & turn on your agent
Important: A test run takes real actions. It will send messages, update files, and set up meetings.
- At the bottom of the agent editor, click Test run
Start.
- This runs the agent immediately so you can verify the output. If everything works fine, you:
- Get a "Run Completed" message.
- Receive a message in Google Chat with a summary of your actual unread emails.
- This runs the agent immediately so you can verify the output. If everything works fine, you:
- When you're finished, click Turn on.
Start.
Explore other ways to create an agent
There are 3 simple ways to create an agent:
- Describe it with AI: This is the fastest way to build an agent. Tell Gemini what you want to automate, and it will build the agent for you. Learn how to create an agent with Gemini AI.
- Use a template: This is the best option for common tasks. Choose a pre-built agent designed for a common task. Learn how to create an agent from a template.
- Start from scratch: For full control, build a custom agent. Choose your own steps and conditions. Learn how to create an agent from scratch.