Google AI Studio: How it works and how to use it (well)

Google AI Studio is a browser-based development platform designed to take full advantage of Gemini AI models. Originally known as MakerSuite, over time it has evolved into a complete suite, thanks also to support for Vibe Coding.

In fact, it's possible to create web applications from scratch by describing via a prompt what you want as the final result. It's based on some very advanced tools, yet its clean interface makes it accessible even to those without specific IT knowledge.

Let's take a detailed look at how Google AI Studio was created, how it works and what it's used for, along with some practical uses and tips for making the most of its potential.

<sommaire|title=What is Google AI Studio and how does it work: the guide>

What is Google AI Studio

Google AI Studio was launched in early 2023 as MakerSuite. Initially, it was a tool designed to allow developers to test Google's language models without having to write complex lines of code.

With the announcement of Gemini, the company decided to rename the tool to Google AI Studio, a platform integrated into the Gemini ecosystem. The interface was changed, becoming more professional to serve as the entry point for those who want to integrate AI into their apps.

The first major breakthrough came with the release of Gemini 1.5 Pro. Here, Google AI Studio became a benchmark for testing a context window of up to 2 million tokens. This allowed users to upload books, hours of videos, and PDF files simultaneously.

Today we've arrived in the era of Vibe Coding, which has transformed AI Studio from a chat environment to a visual development environment. Tools like the Create App feature have been introduced, allowing you to move from text-only to full-featured web applications with code, preview, and deployment.

How Google AI Studio Works

Google AI Studio is actually very simple to use. The basic system works with user-written input. The Context Window is the AI Studio engine, where you can upload files that the model reads instantly. For example, you can upload hour-long videos and ask specific questions, as well as large-scale PDF files.

On the right side of the interface are the parameters you can set. If you lower the "Temperature," the AI becomes more precise and repetitive. If you raise it, it becomes more creative and imaginative. You can also decide how restrictive the safety locks are with the Safety Settings.

After creating the first prompt, you can click Get Code. AI Studio transforms the conversation into pure Python, JavaScript, or cURL code that you can copy and paste into your apps or websites.

A section has been added where you can write what you want to create and the screen splits in two. On the left is the AI writing the code, and on the right is a real-time preview of the app.

What is Google AI Studio used for

Google AI Studio is therefore used to transform an idea into a functioning digital tool, avoiding manual programming. With its context window, it can handle files that would otherwise cripple other chats like the free ChatGPT or standard Gemini.

For those with a simple mini-app idea, Google AI Studio can create the code in a matter of moments. For professional developers, it is a testing ground where they can test prompts, obtain APIs, and compare models.

Finally, it's also useful for those who want to use Google's most powerful models for free, without having to pay a monthly subscription to Gemini Advanced.

How to use Google AI Studio

Want to start using Google AI Studio? First, you should know that the platform is browser-based, accessible from aistudio.google.com. There are no apps for iOS or Android; in fact, the interface is optimized for Windows PCs and Macs.

What are the requirements? All you need is a personal Gmail account or a business Google Workspace account to access, after confirming that you're at least 18 years old. This tool is available in Italy and in most of the rest of the world.

To set up your profile, you must accept the Terms of Service the first time you log in, and then you can create an API Key to test the AI. This is optional, not mandatory.

The interface is easy to understand. In the left column you'll find saved prompts, API key management, and the button to create a new project. In the middle is the spreadsheet where you can write instructions and upload files. On the right is the control panel where you can choose the model and adjust the temperatures.

The free plan can accommodate hundreds of requests per day at no additional cost and provides access to the most powerful models.

How to use Google AI Studio

To move from theory to practice, there are a few things to keep in mind. The first thing to do is not write the question, but set up the System Instructions. Here you define the role and a general explanation.

To upload materials, simply use the + button. You can drag and drop videos, documents, and code folders, with the advantage of analyzing them in a single step.

On the right, you need to configure the template, choosing the most suitable template: Gemini 1.5 Flash (for simple tasks) and Gemini 1.5 Pro (for complex reasoning). For temperature, choose 0.1 for technical data, 0.8 for creative writing.

With the new Create App feature, you can generate a new app from scratch. To do so, describe the application in natural language, and the code will be automatically generated. You'll see the final result appear on the right, so you can test it and request any changes.

After the experiment has worked, you can use the Get Code button in the top right to select the language, copy the code, and paste it into your project. API Key generates the key to make the creation accessible externally.

Some concrete examples of using Google AI Studio

We conclude with some practical examples of using Google AI Studio to take full advantage of its features.

Thanks to multimodality, you can first take advantage of video analysis. Drag and drop a file, even a long one, lasting more than an hour. In the prompt, write: "Analyze this video and create a clickable table of contents with the exact minutes in which marketing budgets are discussed and summarize the decisions made." You'll get the answer in seconds.

For documents, instead, upload dozens of PDF files at once and ask: "Check the termination clause in these contracts. Does any document include different penalties than the others? Create a comparison table."

A third possible use is the Vibe Coding tool for creating a web app. Go to the Create App section and use a prompt like: “Create a web app with a clean interface. It must have three fields: Hourly Cost, Estimated Hours, and Margin %. It must calculate the total and generate text that's ready to copy for the client's email.”

Finally, there's the Code Translator. If you have an old script in a language you're unfamiliar with and want to transfer it to Python, paste the entire file and then write: “You are a software migration expert, convert this code to Python, optimize the functions for speed, and add comments in Italian to explain what each block does.”

Original article published on Money.it Italy. Original title: [Google AI Studio, come funziona e come usarlo (bene)—>https://www.money.it/google-ai-studio-come-funziona-e-come-usarlo-concretamente]