Google AI Studio Guide — Step-by-Step Guide & Transcript

[00:07] an application, test it, improve it, and move it closer to creating a real solution without spending months building everything from scratch. Today, and professionals are moving beyond

[00:19] understanding how solutions are designed, tested, and developed. Now, Google AI Studio provides a space where users can experiment with Gemini models, test ideas, work with different types of information, and understand how

[00:32] moving into the complete development process. In simple words, it allows you to explore the idea, see how it works, make improvements, and gradually move Let's say for example, if you can imagine creating a customer support

[00:47] assistant that can answer common questions, a tool that summarizes large documents, or an application that can analyze images and provide useful starting with a complex development process, Google AI Studio allows you to

[01:01] first experiment with your idea, understand the expected output, and the importance of learning Google AI Studio is not only about understanding a platform. This is about learning how modern digital solutions are planned,

[01:14] beginner exploring technology, a developer building applications, or a professional involving your skills, understanding these platforms helps you solutions. In this tutorial, we will explore how Google AI Studio works, how

[01:29] with different inputs, and how these experiments can be transformed into practical applications. Let's have a look at the agenda for this tutorial. Introduction to Google AI Studio. Understand what Google AI Studio is and

[01:42] how it helps in building and testing applications. Exploring the Google AI platform interface and understand its key features. Creating and testing prompts. Learn how to give instructions and improve responses using practical

[01:55] examples. Working with Gemini models, understand how to select models to test different capabilities. Building practical solutions, explore how ideas tested in AI studio can become real applications. Before we move on, let me

[02:08] guys. If you're interested in understanding how AI applications are actually built and want to move beyond just using AI tools, Advanced Executive Program in Applied Generative AI, which is in collaboration with Simplilearn,

[02:21] can help you develop practical skills for the AI era. Now, this program takes building intelligent AI solutions starting with generative AI foundations and moving into AI agents, rag-powered applications, LLMs, MCP, multi-agent

[02:36] systems, and AI workflow automation. You will also get hands-on exposure to industry-relevant tools and frameworks like LangChain, LangGraph, CrewAI, AutoGen, Claude, Open Claude, Azure AI, and more. Helping you understand how

[02:49] developed. Now, the program is not practical learning through live demonstrations, coding exercises, guided practices, industry use cases, and hands-on labs. You also get to work on

[03:02] real-world AI projects across different domains, including building AI assistants, content generation tools, analytics solutions, multi-agent So, whether you're a developer looking forward to build AI solutions, a

[03:15] technology professional exploring new career opportunities, a business adoption, or someone starting your AI journey, this program helps you gain the skills needed to confidently work with modern AI technologies. Now, the

[03:28] industry tools, expert guidance, and career support makes it a valuable learning experience for anyone who wants to understand how AI systems are created and applied in the real world. So, if your goal is to move from simply using

[03:41] intelligent applications and workflows, this program can help you take that next step. Now, before we continue, here is a quick question for you. What is Google AI Studio mainly used for? Is it A, testing your AI applications, B, editing

[03:55] videos, C, creating websites only, or is it D, managing databases? Let us know now that we know what we're going to explore in this tutorial, let's directly open Google AI Studio and start experimenting with Gemini models. We

[04:07] will begin with a simple prompt, understand how the platform responds, improve our instructions, and then move towards practical examples using text, images, and documents. I'm opening my browser and searching for Google AI

[04:19] Studio. So, from the search results, I will open the official Google AI Studio website. Once the website loads, we can see the Google AI Studio homepage, and this is where we can access Gemini models, create experiments, test

[04:31] different ideas, and later connect these experiments with real applications. I will click on the sign in option available at the top right corner and continue with my Google account. After completing the full login process, we

[04:43] are redirected to the main Google AI Studio workspace. Now, this is where we will spend most of our time throughout this tutorial. Instead of building a complete application immediately, Google AI Studio allows us to first test how we

[04:55] want Gemini to behave. Then, we can experiment with different instructions, check the output, make improvements, and once we're satisfied with the result, connect the same functionality with an application. So, now let me go ahead and

[05:07] click on create prompt to start with our first experiment. Once the prompt workspace opens, we can see the main areas available to us. the model selection options where we can choose the Gemini model that we want to

[05:20] our prompt area where we can provide instructions, and after running the in the output section. Now, let's go ahead and start with a very basic example. Let me go ahead and enter, "Explain machine learning." So, if I go

[05:36] ahead and run this, what you see here is the model processing our request and see that Gemini has explained machine learning, but the answer is quite general. Now, this happened because our instruction was very broad. We only

[05:48] mentioned the topic, but we didn't tell Gemini who the explanation is for and how detailed it should be or what type of example we need. Now, let's go ahead clearing the previous instruction, let me go and enter, "You are a machine

[06:03] me go and enter, "You are a machine learn instructor teaching beginner. learn instructor teaching beginner. Explain machine learning to someone who Explain machine learning to someone who has no technical background." And then

[06:16] has no technical background." And then we will follow that up with "Use a simple real-world example. And then explain the difference between traditional programming machine learning. Keep the explanation under

[06:34] the prompt. Let me go ahead and run this prompt again. As you can see that the explanation that you can see is more structured and beginner-friendly because Gemini the role it should take, the audience that it should explain to, the

[06:49] type of example that we want, and the expected length. Now, I created this Let's quickly understand the components of this prompt. The first line defines the perspective from which it should answer. The second line defines the

[07:03] audience so the explanation is adjusted for beginners. The real-world example requirement here is connecting the concept with something familiar, and the word limit controls how detailed the final response should be. Now, let's go

[07:15] use Google AI Studio for real business scenarios. Imagine that we're building platforms. Now, before creating the complete application, we want to test how Gemini can handle common learner questions. So, I will go ahead and

[07:29] create a new prompt. So, enter, "You are a support assistant for an online learning." And here's what should help the learners with. So, let's go ahead

[07:41] the learners with. So, let's go ahead and type help learners with course information, enrollment questions, certificate related queries, learning certificate related queries, learning support. Keep your responses friendly,

[07:57] support. Keep your responses friendly, clear, and professional. If you do not clear, and professional. If you do not have enough information about then ask them for more details. Now, let

[08:12] me go ahead and click on run. Now, Gemini has created a basic behavior for our learning support assistant. Let's test it with a learner question. So, let test it with a learner question. So, let me go ahead and enter, "I have completed

[08:24] me go ahead and enter, "I have completed my course but have found my certificate. What should I be doing?" So, with this Gemini generates a response based on the

[08:36] instructions that we provided. So, the response is already more focused because we have defined the role, the type of questions it should handle, and more. want to know which course is suitable for somebody starting a career in data

[08:49] analytics. Again, Gemini understands that it needs to respond like a learning support assistant rather than giving a random answer. Now, let's go ahead and improve this assistant by further adding specific information. So, here is the

[09:02] prompt for the same. Additional information, the platform offers beginner, intermediate, courses. Always ask about the

[09:15] courses. Always ask about the background, current skills, and goals before suggesting a course. Recommend learning paths based

[09:27] a course. Recommend learning paths based on the learner's requirement. So, let me go ahead and run this prompt again. So, now the assistant has additional context and can provide more personalized responses. And this is how we build

[09:39] better results inside Google AI Studio. We don't expect the first response to be perfect, but we start with a basic idea, test the output, identify what's missing, and add more information, and gradually improve the behavior. So, now

[09:51] that we have created our refined first text-based experiment, let's move ahead and explore one of the most useful capabilities of Google AI Studio. And that is working with images and understanding visual information. Until

[10:03] now, we have seen how AI can respond to text-based instructions, but modern AI systems are not limited to only understanding words. They can also analyze images, identify objects, understand visual details, and provide

[10:15] meaningful responses based on what they see. So, this capability is useful in many real-world scenarios. Say, for example, an e-commerce company can analyze product images and automatically create descriptions, a support team can

[10:28] understand screenshots shared by customers, and an education platform can use AI to explain diagrams, charts, or learning materials. Now, let's see how this works inside Google AI Studio. So, I will start by using the image upload

[10:40] option available in the prompt area. As you can see, you have this option over here on the right of the prompt box. You will have to go click on this, and here you have an option to upload files. So, for this particular demo, I'm uploading

[10:53] a smart watch product image. So, once the image is uploaded, Google AI Studio with the instructions that we provide. So, instead of simply asking the AI, detailed instruction to get a useful response. So, I will be using this

[11:08] response. So, I will be using this prompt. Analyze the smart watch product prompt. Analyze the smart watch product image. Provide a detailed description image. Provide a detailed description of the product, important features

[11:22] of the product, important features available or visible in the image, targeted customers who may be interested in this product.

[11:35] who may be interested in this product. Three marketing ideas Three marketing ideas for promoting this product. So, let's go ahead and run this prompt. So, the prompt is structured into different

[11:47] parts. So, first we tell the model what we want it to analyze. Next, we define the type of information that we need from the image. Finally, we specify the output that we expect, which helps the response become more focused and useful.

[11:59] see, Google AI Studio is able to identify the smart watch from the image, understand the details visible in the picture, and generate insights based on the instructions. The important thing to understand here is that AI is not just

[12:12] the visual information with our instructions and generating a response based on the context that we provide. It can identify that the image contains a smart watch, understand that it's a wearable technology product, and suggest

[12:24] possible customers or marketing approaches based on that understanding. Now, this ability to understand visual information can be applied across different industries. Businesses can use it for product analysis, designers can

[12:36] support teams can use it for troubleshooting screenshots, and complex images or diagrams. Let's take can combine images with additional text instructions to create even more useful

[12:50] results. So, now that we have seen how Google AI Studio can understand visual information, let's take this a step further and see how it can work with multiple types of inputs together. So, in real-world applications, information

[13:02] is rarely available only in one format. For example, a business might have product images along with descriptions, a student might have handwritten notes with questions, or a company might have screenshots along with customer issues.

[13:15] capabilities become useful because the model can understand the relationship information. Now, let's go ahead and test this by uploading an image and providing additional instructions. So, at first, let me go ahead and click on

[13:28] the image upload option. So, now instead of simply asking, "What is this image?" I will provide a more specific instruction. Let's say, for example, so this prompt is divided into three parts. First, we're giving the model a task,

[13:40] which is analyzing the image. Second, we're defining the output format. Third, we're guiding the model on what type of information we need. So, as you can see, Google AI Studio is not only identifying objects in the image, but also

[13:52] useful insights based on the instructions. Now, the same approach can be applied in many scenarios, like product analysis, document review, educational tools, design feedback, and customer support applications. So, now

[14:05] with multiple inputs, let's move on to another important capability, which is working with documents and files. Now, let's explore how Google AI Studio can help us understand large amounts of information from documents. Now, imagine

[14:19] you have a research paper, company policy documents, product manuals, or a reading everything, you can provide the document and ask the specific questions. Now, let's upload a sample PDF file. After uploading the document, I can ask

[14:33] questions such as, "Summarize this document in simple terms." So, let me go ahead and type that out. "Summarize the document attached in simple words." So,

[14:46] let me go and attach the file. "Highlight the key points. Create a list "Highlight the key points. Create a list Sorry, list of important action items."

[15:00] So, here summarize tells the model what action to perform, and simple terms defines the style of response. Here, key expected output. Now, let's go ahead and execute this. So, the important thing to

[15:14] just finding keywords. It's also understanding the information inside the document and generating responses based on context. So, this capability can be business teams, and professionals who want to regularly work with large

[15:30] information. So, now that we have explored text, images, and documents, improve responses by giving better instructions. So, now that we have seen different inputs, let's understand how the quality of output depends on how

[15:44] clearly we communicate our requirements. Say, for example, instead of writing explain marketing, we can write explain digital marketing to a beginner. Use

[15:56] digital marketing to a beginner. Use simple examples. Explain in five bullet points. Yeah. So, when I go ahead and run this, we can tell how the improvement comes from adding the current audience, the expected format,

[16:09] the level of explanation, and so on. So, now let's go ahead and compare both outputs. The second prompt gives us a much focused answer because we provided important skill when working with AI tools because better instructions help

[16:22] go ahead and see how Google AI Studio can help developers and beginners create code. For example, I will ask Gemini create a simple page for a personal

[16:35] create a simple page for a personal portfolio website with a navigation portfolio website with a navigation bar, profile section, and contact form. So, this particular code contains an HTML structure for the web page, CSS

[16:49] styling for appearance, sections required for the portfolio page. So, now let me go ahead and open Visual Code Studio. So, as you can see, I've opened Visual Studio Code and created a new HTML file. So, I'll go ahead and paste

[17:01] browser. So, as you can see, I've copied the code. Now, let me go ahead and paste in the file explorer, so as you can see, Gemini has helped us create the basic replace understanding programming concepts, but it helps developers and

[17:16] beginners build, test, and improve ideas faster. Now that you have explored Google AI Studio, you have a better understanding of how ideas can be practical applications. From experimenting with Gemini models to

[17:29] problems, this platform provides a great starting point for anyone interested in building with modern technology. Keep exploring, keep learning, and start Subscribe to Simply Learn for more such insights and check out our related

[17:43] insights and check out our related videos.