Gemini Code: NEW Agentic AI Coder By Google! Automate Coding! (Installation Guide)

When cloud 3. 7 sonnet released the enthropic team had released a thing called claude code which is a command line tool for gentic coding. Cloud code can search and read code edit files write and even run test. It's something that can even commit and push code to GitHub which is making it a versatile AI coding tool within your command line keeping you in the loop each step of the way. Now Claude Code's quality is quite impressive. However, the pricing is kind of inefficient for anyone to actually use in any project because the input token as well as the output token is quite pricey. And you can see that it might not even be viable to use in a long-term project. But don't you worry because there's another capable model with similar coding performance that also comes with a command line tool for gentic coding. Allow me to introduce Gemini code. This is an AI powered coding assistant that brings the full strength of Google Gemini 2. 5 Pro directly within your terminal. We all know that this is the state-of-the-art model currently that ranks number one across multiple benchmarks out competing many of the models that we all know and love like OpenAI's 03 Mini, GPT 4. 5, Claude 3. 7 Sonnet and many others. You can see in almost every category it is packing a punch in terms of its performance and in particularly with coding. It actually does a really great job in almost every benchmark. Now I know some of you guys might say Gemini 2. 5 is kind of an experimental. So the request per day limit is kind of something that might hinder your usability with this because you get 25 requests per day while tier one bumps you all the way to 100. So what's the solution to that? Well, you can obviously fall back on many of the other Gemini models like 1. 5 Pro, which I will be honest is a bit clunky right now, but it's going to get the job done and it's kind of efficient in terms of generating code. So, you can definitely choose between these two options to get the job done. With Gemini code though, you can do practically everything that you could do with Claude Code. You have an interactive chat session within your terminal, meaning that you can start interacting with these models within your command line. You have multiple support for different models that Gemini provides like the 2. 5 Pro, 1. 5 Pro, and many others. You have a history management system, markdown rendering within the terminal. You have automatic tool usage by the assistant, which is going to be really helpful cuz this model has a great job in terms of tool use. You have file operation where you can view edit list as well as GP and Glob. You have directory operations. You have system commands, quality checks as well as test running capabilities. Now to get started, there's two methods. You can install this from Pippi or the source where you can clone the repository and go through and install the packages. There's super simple steps to get started. So I'll simply go ahead and showcase this. If you are going to be following through with my method, I'm going to be using pipi. So, let's go over to our command prompt and copy that command. It's super simple. Just go ahead and paste this in and you're going to be able to then go ahead and install everything that is necessary to have Gemini code functional on your computer. So, once it finishes installing, we can then easily go ahead and set this up. So, you can set it up with your Google API. This is with the ability to use different models like Gemini 2. 5 as well as other models. Now, to get your API key, you will need to go over to Google AI Studio, which will require you to have a Google account. So, make sure you have that set up and click on get API key. This is where you can simply go ahead and create an API key and then make sure you have it linked to a billing account. Otherwise, this is not going to work. It is currently free and you have 25 requests for free if you're on tier one. And obviously two, you have 100 free requests, which is so many different requests that you can execute locally. So what you'll need to do is simply go ahead and copy Gemini setup, paste it into your command prompt, and then paste in your API key after you have pasted in a space and click enter, and it will state Gemini API key has been saved. Now, what you can do is before you actually start using it, you're going to need to go ahead and start the interactive session with the default model. But if you want to use the specific 2. 5 Pro model, you will need to use this command. So, make sure you go ahead and set that as your default uh model first, and then you can go ahead and start the session with it. So, in this case, I'm going to go ahead and paste this in to set the model. And once that is set, we're going to go ahead and start the session with the Gemini 2. 5 Pro. And that's by simply facing in this model. And you're going to be able to see that Gemini code is now fully started up. Now, if you actually need help with anything, you can obviously press help. And it'll tell

you all the different uh commands that you can use such as the bash, create directory, edit, and whatnot. But let's go ahead and test it out. And let's generate something. We can say go ahead and create me a SAS landing page for World of AI YouTube channel. And then we can go ahead and click enter. You'll see that the assistant is going to be thinking cuz this is a thinking model. And it's going to go ahead and plan through the steps with the AI agents. And all of these actions will be executed within your terminal, which is just amazing cuz you can see now this is going to be a great way for you to easily execute, build, as well as debug right within your terminal. And what makes this even better is not using it with this command prompt. It's better if you have it integrated within your IDE's terminal, which makes the whole process easier. So, it looks like it has created the directory. We're going to go ahead and take a look at it. And there we go. But with a single prompt, you can see that it was capable of creating this beautiful SAS website. This is a nice landing page and it was capable of doing it within my terminal. This is how impressive this model is, especially with this command line tool. Now, particularly it's not going to be used to fully code out everything in terms of building application, but it's going to be an assistant tool that can help you in terms of refactoring as well as debugging certain snippets of code. Now, just keep in mind this is just an independent tool that was built as a hackathon project. So, it's likely not going to get a lot of edits in the future. It is built by some developer on Twitter, so keep that in mind. It's not an official project by Google. But anyways, I'll leave all these links in the description below so that you can easily get started. There's a lot more information on their GitHub repository as well as the developers Twitter. So, definitely give them a follow. But with that thought guys, I hope you enjoyed today's video and got some sort of value out of it. Follow me on the newsletter, Patreon, as well as follow me on Twitter. And make sure you guys subscribe, turn on the notification bell, like this video, and please take a look at our previous videos cuz there's a lot of content. Well, not a lot actually cuz you only have one video, but I guess you have a lot more content on the main channel if you haven't seen already. But I'll leave a link to that in the description below so that you can take a look at it. But with that thought, guys, have an amazing day. Spread positivity and I'll see you guys fairly shortly. He suffers.