How to Use Gemini 2.5 Flash in Apps Script with Vertex AI

Key Takeaways

In this video, I'm going to show you how to use our Vertex AI advanced service to prompt the Gemini 2.5 flash model directly from an Appcript project. But first, what is Vertex AI and what makes it interesting to AppScript and Google Workspace developers? Google Cloud's Vertex AI is a machine learning platform that lets you train and deploy ML models and AI applications. There are a few reasons why you might be interested in using Vertex AI as an appcript developer. First, the Vertex AI's model garden provides Appcript developers with enterprisegrade access to Gemini and thirdparty models. The platform also supports advanced agentic features like function calling and fine-tuning while offering enhanced security and access control. These models and features ultimately make Vert.exai the preferred choice for scalable professional Google Workspace solutions that you can share with others in your organization or publicly on the Google Workspace marketplace. To use AppScript's Vertex AI advanced service, you need to have a Google Cloud project. In the cloud project, you must also enable billing and the Vert.ex AI API. Now that we have what we need, let's set up an Appcript project with Vert.Exai. If you'd like to try this out yourself, the instructions and code sample I'll be using are available in the AppScript documentation, which I'll link in the video description. Okay, so here's my Appcript project, which has some sample code. I'll start by enabling the Vertex AI service. From the editor, I click add a service. And then from the drop-own menu, I select Vertex AI API. After I click add, you can see that Vertex AI now appears under my enabled services. Now, let's look at the code. First, we're using the main function as the entry point to trigger the advanced service. This function includes the prompt we'll use and handles any errors if the connection to Vertex AI fails. Our prompt asks what is AppScript in one sentence. The main function triggers the function callver vertex AI which handles the API request by specifying an AI model and the cloud project ID and region. I've set my cloud project ID here as a script property, but you can also just specify the ID as a string. Here's where we specify the Gemini 2.5 flash model. This is the model we'll prompt with our question about AppScript. And here's the payload for the request to Vert.ex AI API, which includes the prompt and some other configurations, including the maximum output tokens for the response. In this line, you can see we're using the Vert.Ex AI advanced service to call the API's generate content method. The advanced service acts as a built-in authenticated wrapper that sends your request directly to the Vert.Ex AI API. To test this code, I go to the top menu and click run to run the main function. A window may pop up asking you to give permissions for the function to run, but otherwise it'll run right away as you see here. And in the execution log, the service has answered our prompt and generated some text explaining what Appcript is. And that's all there is to getting started with Vert.Exai. To try it out yourself, check out our quick start in the Appcript documentation. The quick start shows you how to configure your Google Cloud and Appcript projects and includes the code sample that I demoed for you today. For more advanced examples showing you how to use Vertex AI to build AI agents and add-ons for Google Workspace, check out our samples gallery. Thanks for watching.