OpenAI Codex Full Course 2026 | Learn OpenAI Codex From Scratch | OpenAI Codex Course | Simplilearn
Chapters6
Defines CodeEx as a project aware AI agent that reads codebases, edits files, runs commands, and helps manage development tasks beyond simple code snippets.
Simplilearn’s OpenAI Codex Full Course teaches Codex as an AI coding agent—across CLI, IDE, web, and app—and walks you through building a real project end-to-end.
Summary
In this Simplilearn course, the instructor introduces OpenAI Codex as more than a code snippet generator—it's an AI coding agent that can read an entire project, make changes, run commands, and help you ship features. You’ll see Codex used across four surfaces: CLI, IDE extensions, the web, and a dedicated app, with practical demonstrations of installing, configuring, and prompting Codex for real work. The walkthrough explains how Codex differs from ChatGPT (project-first vs. conversation-first) and covers prompting strategies, debugging, code review, documentation, and testing within an active workflow. A central demonstration builds a web-based face-transformation MVP, showing how Codex can bootstrap a project, generate files like agents.md, wire in a design layout from a Pinterest reference, and perform web searches to enrich documentation. The video emphasizes best practices, such as starting with understanding the project, using version control, and writing precise prompts, rather than asking Codex to do everything at once. You’ll also learn how to switch between models, select reasoning depth, resume sessions, and use fast mode for speed—including how to integrate with GitHub and deploy to various surfaces. The course weaves in practical tips about IDE integration (VS Code, Cursor), cloud tooling, and the desktop Codex app, plus a final reminder that Codex accelerates work but still requires developer oversight. If you’re aiming to turn generative AI into a productive coding workflow, this beginner-friendly course covers the fundamentals and practical workflows to get you building quickly.
Key Takeaways
- Codex operates as an AI coding agent that can inspect a project, edit files, run commands, and generate documentation and tests, not just provide code snippets.
- You can use Codex via CLI, IDE extension, web, or the Codex app, with a common workflow of inspecting, debugging, and building features inside your real repository.
- Resume and new-session features allow picking up work where you left off or starting fresh while preserving context for long-running tasks.
- Prompting matters: specify task context, constraints, validation, and expected output to guide Codex effectively and avoid vague, unfocused results.
- Model and reasoning level selection (GPT-5.4 high vs. other modes) impact the depth and speed of Codex’s reasoning and code generation.
- Best practices include starting from project understanding, using Git for checkpoints, writing precise prompts, and testing changes before shipping.
- Codeex enables end-to-end workflows—from planning to deployment—across CLI, IDE, web, and desktop app, supporting collaborative and scalable software development.
Who Is This For?
Essential viewing for developers and teams who want to accelerate coding with AI agents. Beginners will learn the basics, while experienced developers will discover practical workflows to move faster across debugging, testing, documentation, and feature work.
Notable Quotes
"Codeex is an AI coding agent. It can read your project, understand files, make code changes, run commands, review code, explain unfamiliar code bases, and even help you create pull requests."
—Defines Codex’s capabilities beyond simple code generation and sets the stage for project-wide automation.
"Codeex is designed to work across your actual project files. That means instead of only giving you an answer in chat, Codeex can inspect your code base, understand your folder structure, edit files, run commands, and help move the work forward."
—Highlights the “project-first” vs. “conversation-first” distinction between Codex and ChatGPT.
"The very first thing you want Codeex to do is to create agents.md file. So I’m going to give the prompt create an agents.md file in this project and keep it empty for now."
—Shows a practical, repeatable first-step workflow when starting a Codex-powered project.
"The future of coding is not just writing every line manually. It is knowing how to guide AI agents, review their work, and build better software faster."
—Capsules the video’s takeaway about how Codex fits into modern software teams and productivity.
"This tool is incredibly powerful because you can use it for marketing, for hiring, for growing your business in all kinds of various ways for managing your personal productivity."
—Emphasizes Codex’s versatility beyond pure coding tasks.
Questions This Video Answers
- How can OpenAI Codex be used across CLI, IDE, web, and app surfaces?
- What are best practices for prompting Codex to build a feature end-to-end in a real project?
- How do you resume a Codex session or switch to fast/plan modes during a project?
- What are the differences between Codex and ChatGPT for software development?
- Which tools and workflows are recommended to integrate Codex with GitHub and deployment pipelines?
OpenAI CodexCodex CLICodex IDE extensionCodex WebAgentic AIAI coding assistantPrompt engineeringSoftware development workflowsGitHub integrationNext.js and deployment
Full Transcript
Have you ever opened a code base and felt completely lost? Hundreds of files, confusing folder names, bugs, hiding somewhere, and one small change breaking five different things. Now imagine this. Instead of manually searching every file, writing every function from scratch, and debugging line by line, you simply say, "Codeex, understand this project. Find the bug, fix it, add test and show me the changes. That is the idea behind OpenAI codeex. It is not just a chatbot that gives you code snippets but codeex is an AI coding agent. It can read your project, understand files, make code changes, run commands, review code, explain unfamiliar code bases, and even help you create pull requests.
So in this full course we are going to understand codeex from zero. By the end of this video you will know what codeex is, how it is different from chat GPT, how to use it through CLI, IDE, web or app, how to prompt it properly, how to use it for debugging, building features, code review, documentation, testing and real project workflows. So before diving into video, let's look what exactly we are going to cover in this video. So here first of all we will cover fundamentals of codeex then installing codeex setting up the environment and understanding how it works within ID.
Then we will build a real app together from scratch. After that we will go into advanced features like sub agent skills automation MCP servers cloud agents and more. And finally best practices and mistakes to avoid when using codeex in real projects. So this course is beginner friendly but we will also cover practical developer workflows. So even if you already know coding you will still find this useful. Also just a quick information simply learn in collaboration with IITM Parvat Technologies Foundation brings you the advanced executive program in applied generative AI. This 4 month live online program help you master generative AI, LLMs, prompt engineering, rack, agentic AI and MCP frameworks.
You will learn through 70 plus hours of expertled sessions, 60 plus hands-on exercise and seven industry related projects. The program also gives you exposure to the tools like chat GPT, Azure AI studio, copilot, hugging face, tally, lang chain and much more. Learners also get exclusive master classes by IITM and IIT faculty or aluminina. You can become eligible for a two-day campus immersion at IIT Madras research park. On completion, you will receive a program certificate from IIT Technologies Foundation. You also get career support through simply learns job assist to help you get noticed by top hiring companies.
So if you want to build real world generative AI skills, this program is a strong place to start. Do check out the link in the description box and in the pinned comments for more details. So before we deep dive, here's a quick quiz question for you. And the question is, what is the main purpose of OpenAI codeex? And your options are option A to edit videos automatically, option B to help write debug and review code, option C to create social media posts only, or the option D to store files in the cloud. Let me know your answers in the comment section below and let's see who will give the right answer first.
So without any further ado, let's get started. So let's first understand codeex in a very simple way. See, JPT is great when you want to ask questions, brainstorm ideas, understand concepts or generate code in conversation. But Codeex is designed for different kind of work. Codeex is built to work across your actual project files. That means instead of only giving you an answer in chat, Codeex can inspect your code base, understand your project structure, edit files, run commands, and help move the work forward. OpenAI describes Codeex as a AI agent that helps you write, review, and ship code.
Think of Codeex like an AI software engineering assistant. You can ask it to explain how your project works. You can ask it to find a bug in an API. You can ask it to add authentication to an app. You can even ask it to write unit test for a function. You can ask it to review a pull request. You can ask it to create documentation for a module and you can even ask it to modernize and old code base. So the important thing to understand is codeex is not only generating code. It is helping you work with a real code base that makes it useful for developers, students, software engineers, product teams, startup founders and even non-developers who want to understand or manage technical work.
Now let's answer a very common question. If chair GPT can already write code, why do we need codeex? The simplest answer is this. Chair GPT helps you think through the work while codeex help you hand off parts of the work. For example, if you ask chair GPT to write a login form in React, it can give you a code snippet. But Codeex can do something more practical. You can open your existing React project and say add a login page that matches the current design system, connect with the existing API service, validate email and password, handles loading state and add tests.
Now Codeex can inspect your project. It can understand your folder structure. It can follow your code style. It can modify the right files and it can show you the changes. That is the difference. Chair GPT is mostly conversation first but Codeex is project first. Chair GPT gives answers. Codeex can take action. Chair GPT can explain code. Codex can explain, edit, test and review code inside your real workflow. So this does not mean CEX replaces developers. It means CEX helps developer move faster. You still need to understand the requirement, review the output, test the final result and make decisions.
Codeex is powerful but you are still responsible for the final code. So basically we have understood what codeex is and you can even see on the screen I have opened the official document of codeex where it has explained what codeex is, what all features it contains, what all task it performs. So you can visit this official docs of codeex and get all the information out there. I highly recommend you checking out this documentation as it has everything you need to know. So let's start by understanding where you can use codeex. So codeex is available across different development service surfaces.
So first of all you can use it in the terminal through codeex cli. See here here under using codeex you can see cla. So CLI stands for command line interface. So one of the ways of using codeex is using codec cla. Another thing you can use inside your editor through the codeex IDE extension. Another way of using codeex you can visit here then clicking on overview and you will find how we can use codeex in our IDE. It has complete information and guide what all you need to know, what all you need to set up the entire steps and information.
Here we have features, settings, ID, commands, slash commands and whatnot. Now you can even use it in the cloud through codeex web. So another way of using is here you can click on web and you can click on the overview and get the complete information about codeex web. It means you can use codeex in the cloud. So Codeex is a openi coding agent which you can connect with your GitHub account and you can use on the web. Learn about prompting common workflows configuring environments delegate work from the ID extension delegating from GitHub control internet access and whatn not and last way to use it through the codeex app.
So this is something amazing and a newly added feature of using codeex. So you can see open AI codeex app docs mention that codeex can be used through interfaces. So you can visit this codeex app. It is focused on desktop experience for working on codeex threads in parallel. So these are the steps how you can start your codeex app in cloud and in the desktop app. So follow these steps from these official document. So according to open AI docs, it has been shown that codecs can be used through interfaces like ID, CLI, web, mobile and CI/CD workflows.
So now let's understand all these four ways in a little detail and let's again start with CLI. So codec is used from the terminal. This is useful when you want codeex to work directly inside your project folder. Codeex CLI can read, change and run code on your machine in the selected directory. A typical workflow looks like this. You open your project folder. You run codeex. You ask it to inspect the project. Then you ask it to find the bug, build a feature or write tests. For example, you can ask Codeex to explain the project structure.
You can ask it to find why test is failing. You can ask it to add error handling into the payment API. You can even ask it to refactor a file without changing its behavior. And this is very useful for developers who are comfortable with terminals. So this was the basic idea of working with codec cla. Now let's understand about this in little detail. So the codeex ID extension is useful when you want codec inside your code editor. For example, if you use VS code or a similar editor, you can work with codeex without leaving your coding environment.
You can select code, ask questions, request changes and review modifications. This is useful when you are actively developing and want codeex beside your code. Now talking about Codex web. So Codex web lets you delegate task to Codex in the cloud. Codex can read, edit and run code. Codex cloud can also work on task in the background including in parallel using its own cloud environment. This is useful when you want codeex to work on a repository while continue doing something else. For example, you can ask codeex to fix failing test in your repository. You can ask it to add documentation for an API module.
You can ask it to upgrade a dependency and handle breaking changes. You can ask it to review a pull request. So, Codeex does all this for you. Codeex can even work in its own environment and then shows you the proposed changes. And now the last one is Codeex app which I have explained you earlier but let's revise. So you can use any of these four methods to use codeex whatever suits your requirement. So let's look again at the app of codeex. So let me click on the overview part. You can scroll down and explore all its features.
So it is useful when you want a dedicated workspace for agentic coding. It works like a command center for agentic coding with work trees and cloud environments for parallel work. The app helps when you want to manage multiple coding tasks, review diffs and continue work across different surfaces. So the simple explanation is this. Use CLI when you want terminal based control. Use ID when you want codeex inside your editor. Use web when you want cloud-based tasks. And use the app when you want command center for multiple coding workflows. So choose according to your requirement and start building something amazing.
Now I'll show you which is the one that absolutely anybody can use it no matter how bad your computer is. And so that is codec. So open your codec in the official documentation. When you will scroll down you will find these three commands to install, run and upgrade. So it is very simple to install CLI by just copy pasting this command. Install the codec cla with npm. Once you will copy paste it in your terminal, you are good to go with codeex. So let's copy this command here. Let's enter into command prompt and paste it here and press enter.
This particular command will install OpenAI codeex in your device. See in just 5 seconds it's all done. Now just type codeex high-en version to verify the installation. So write it codeexy version and press enter. And this verifies your installation that it is successfully installed in your device. Now let's get into codeex and start building something mindblowing. So for this just type codeex on your terminal and press enter and you are inside codeex. So when you will do this it will ask you to sign up with your account. So whatever chat GPT account you have you sign up with this and another most important thing I want to explain you is this that go with a paid plan because does not come with a free plan of chat GPT.
It is a paid plan. So let me show you about the codeex pricing. So basically we have free go plus and pro plan. So in this free plan you will not get enough number of credits to work on codec. So I'll highly recommend you to go with either go or plus. So the most recommended is plus plan and it is not that expensive just $20 per month and you're good to go and you can build less of application by just investing few dollars. Here you can see codeex on web in the CLI in the ID extension and on iOS.
So this plan offers you the working of codec. So go purchase this. It's highly recommended. You will not regret trust me. No worries. If you don't have any of the plans, you have still two more options to use Codeex. First is signing up with your device code or API or the second one is using the API key. Now, honestly, device code only works if you have another device that is signed in. So, really the only two relevant options are signing in with your chat GPT subscription or the API key. Now since most of the people already have CHGP subscription and I also hope that you are also having one.
So I'm going to show you with this. So go back to your terminal and connect it and sign it with your GPT account your paid plan basically. And here I have been successfully connected where my model shows as GPT 5.4. Now let's see it's working whether it is working fine or not. For this we'll type hey. It is responding hey. So it is successfully working and it is connected with RGBT. Let's type who are you? So it says I am Codeex your coding agent in this workspace. I can inspect the codebase, edit files, run commands, debug issues and explain what I changed.
So you can easily chitchat with your codeex platform. So there is further more than chitchatting. Let me show you this. So, Codeex is a competitor of cloud code, but Codeex without a doubt is more powerful at building complex applications, solving deep errors, and running for longer without making mistakes. So, if you aren't using Codeex in 2026, you really are missing out because Cloud Code while it is good, it's not as powerful as Codeex at many of the use cases. And by the way, if you're having any troubles with the codeex to install or it's or if it's installation doesn't work or throwing some errors, then you might need to check if you have NodeJS installed in your system properly.
If not installed properly, you can visit its official website and you can download it from there and you can type node-v to check if it is properly installed or not. And because it was properly installed in my system, that is why we have faced no issues in installing Codeex. So before installing Codeex, make sure NodeJS is Now let me show you how we can select models. Here you can see I'm already into GPT 5.4 high model. But by pressing /mod, you can choose different models for your device connection that is GPD 5.4 4, GPT 5.5, GPT 5.4 mini, codecs and GPT 5.2.
So these are different models available but I highly recommend you choosing the latest model of GBT that is GBD 5.4 because it shows current strong model for everyday coding. So enter this here you will select the reasoning level for your GBD 5.4. So there are basically four reasoning level which is low and it is better for fast responses with lightning reasoning. Second is medium which is by default. It balances speed and reasoning depth for everyday task. Now the third reasoning level is high which we are into current because it shows the current. We are into the high model of GPT.
So it is greater reasoning depth for complex problems because we are going to make complex projects in our device. That is why I have recommended and chosen high for my project. And here comes the extra high version. So the low version I would never suggest you to use. This is simple too because the simple reason is it has too little reasoning and it will underperform very badly. So personally I would never use low and most of the time I do is set it to high actually because the beauty of high is that it if it needs to it can run for 2 3 4 minutes if the task requires it.
I think high is the sweet spot and that's what I'm going to select here. So model change to GBD 5.4 high. Now let's give the prompt as where on the system are you running. So just to check where it is actually running in our device. So where in the system you are actually running and press enter. Let's see what it gives. So it shows my current working directory is C drive and under the user SLP15185. So this is the current location where our codeex is working. So this is basically our root level and it would be good running it on the root level but if you want to build something like an application you know or do some AI coding it's much better to launch it in a specific folder but as of now I am running it inside this directory itself.
Now we can press Ctrl C to kill this current process. Ctrl C. So it has killed our present session and here we need to change the directory now and we want to build a project inside the documents. So how we can do this like this cd that is stands for change directory cd documents and press enter. It will take us inside the documents folder. Now if you want to go out of this document you can simply press cd dot dot and it will take out from the documents folder. See we're out. And if you want to clear something you can simply press cls and it will clear all our previous history and bring a fresh new terminal.
So if you want to check where we are currently so we can directly press cd enter and it will show we are not inside any of the folders we are just inside the major root level. Now the another important command is diir. You must be thinking what it does then it shows us all the files and folders along with sizes. So here you can see we have all the files, all the folders, the date and the size. Now we want to work now let's say we want to work inside a proper project folder. So here we will make a new directory and the command to do the same is mkdir that is useful in making directory codeex project.
This is the name of my directory. Make sure if you are giving two names like two words with a space then put the name inside double quotes or if you're giving just a single word as the name of your directory let's suppose codeex or just project you can use without double quotes but here I'm using the two words that is why I am closing them inside a double quote so it will consider it as a single name as codeex project and press enter now we will go inside this directory. Here we have just created the directory.
Now press cd codeex project to go inside the codeex project and start our project. Actually enter. See we are inside the codeex project now. Now we can finally launch codeex in this particular folder. So press codeex. Enter codeex. Press enter. So do you trust the content? I say yes. Continue. and we are all good to go. Another thing I want to make you understand is don't be one of those people who saves everything on the desktop. That is not a good way of doing work. You should avoid that at all the cost. Just have a specific folder for each coding project you are going to build and you are good to go there.
Now the very first thing you want codeex to do is no matter what you are building is to create agents.md file. So I'm going to give the prompt create an agents.md file in this project and keep it empty for now. Just put a header. This is the ultimate system prompt that all the agents should follow. And I pressed enter and it has started working on this. So it has created agents.m MD file for us just with an header. And here's the location where it has created the MD file. Now I'll show you what should be the text the format of the text that should be stored inside agents.md.
And here we have the complete prompt. Let me copy paste this prompt. Let's go back to our terminal and let's say put it in the following text inside agents dot MD and let's paste our prompt. So it is saying I'm replacing the placeholder header in this file with the template you have given. Let's wait for it and it's updated with the exact template which we have given above. This is already a very decent agent MD file that we are going to make as agent behave better, respond in a more friendly way and follow best practices.
So need not worry I'll attach the link to this document uh with the description box. It is completely free. Take it, use it and build your project. As of now we have just seen the basic commands of using codeex. Now we will build a project from scratch. A project where we will use all the commands and all the features of codeex. And the project we are building here is a face transformation app. So what we are building now is the app for women targeting apps. The instinct is to build something for yourself for the women community.
you know as you want to build a tool that you would use and you can go into this highly competitive industry but there's so much potential building these mobile apps but here as of now we are building the web app for the same so actually I have really good idea of showing how we would look like this app to be so the basic idea is AI images are much newer especially since the rise of nano banana and Nano Banana 2. Then the actual AI text, right? Most people understand AI text. So what we are going to do is we are going to use the potential of AI images to build an app.
And this is something that literally every single women or girl is wondering how she would look if if she had a bit smaller nose or if she had a bit lip filler or stuff like that, right? So I think it's a good idea from the technical standpoint because I'm going to because it's going to allow us to use AI images and it is a pretty interesting project to build. So whatever changes you want to do in your face like you are going for a plastic surgery and you want to see how the particular changes would look like on your face whether you are just filling your lips or doing a botox or something like that you can directly ask them make the changes on my face and if it looks good you can really apply it in your real life and you're good to go for this decision.
So I think this is really useful app and many women and many people will like this. So let's start building this amazing project. So I'm not going to build this project. I'm just going to give my idea in the prompt format to the codeex and codeex will build it on my behalf. Let's wait and watch. So now I'm here giving a prompt like this. I want to build an app where you can upload an image of yourself and see how would you look like with different facial changes, smaller nose, lip filter, foxy eye lift, face lift, Botox, different skin tone and at least 10 plus other things that women often do or would like to see like when would look on them before doing anything.
The target audience is women and the limit is really between 18 to 55 years old. So I am sure that I have given everything or as much as the codeex wants the idea to build an app. Now let's press enter and see what codeex has to present us. Now I'm g update agents.md and readme.md files. For this it is very important to update everything in these files. We have pressed enter and codeex have started working on it. So the prompt is like it's updated already and we can reflect the changes and see the changes in both the files.
I will also show you how we can connect our codeex account with the GitHub and it is very essential because it can save our code there and whenever we want or we want to share it with any other person we can directly share our GitHub repository with them. So let's go to the Pinterest and from here we will select the layout of our app. How we want our app to be looked like. So scroll down. Select any of the layout according to your own choice. Uh let's suppose I like this. So I will copy the link address of this Pinterest layout.
I will go back to my editor and I will press Ctrl V and then I'll give the chrome test. This is what the layout of the app should look like after the user uploads his or her face. It should generate all kinds of different predefined changes to her face. Again, at least 15 to 20 different images, study this layout, and save it into the MD files. Now, another thing I'm going to show you is web search. This is one of the best built-in tools into Codeex, but again, very underutilized. So, I'm going to give another prompt saying, "Use the web search tool to find the 20 plus most common facial cosmetic surgeries women do and save them into a single bullet list into a readme.md file.
We are using web search for more information and more accurate results. here. So let me give the prompt as I've said this. So I have given the prompt here for the web search tool and then we will press enter. So it has started working and generating the desired output for us. See meanwhile it is working I want to clear one of your doubts. See codeex is actually less of coding and more of marketing. I'm learning more about that my target avatar for this built-in project to make the app better, right? And this is something you can do when writing a landing page, when writing email copy.
Use these AI agents to understand the industry more, to understand your users more, to write better copy, to do competitive research. There's so much you can do with the tool called Codeex. See people think it's only for coding because it's in the name right codeex but it's not. You can absolutely and exactly use it for marketing, for hiring, for growing your business in all kinds of various ways for managing your personal productivity. And there's a literally endless ways you can use codeex. Anything that you as a human can do on computer, you can use AI agents like codeex to help you save your time and even do it fully automatically.
So meanwhile I was giving and sharing some information. It has performed his search. So here Codex did multiple different web searches until it had enough information and then it has saved the results into our files that is much more accurate than what I could even expect. So it is great and I hope this will bring a lot of accuracy in our project. Here I want to show you another cool feature of codeex. Let's say you accidentally shut it off or maybe you don't use the computer for a few days and then you want to go back to the previous session but you already know whenever we use codeex it starts with a new session right but guess what there is a command called / resume where you can resume your saved chat and start your coding from there where you left so let me show it to you practically let's suppose we have cleared everything.
Then we run Codeex again. After a few days, let's suppose we are running Codeex again and we want to continue building our app because of some reason we left it somewhere. So we want to resume. So we'll type resume. This command helps us resume the saved chat. So press enter and it has get us back to the project where we have left. So it is showing 12 hour ago I was working on this project and this is where I seriously want to land again. So I'll place I press enter and here we are back again in our folder in our project which we were building that is face transformation app.
Now we have like we have conducted all the searches all the data is there with codeex now but now I want the codeex only to suggest me some names. So I'll give the prompt accordingly. Suggest at least 10 at least 10 names for my project. I don't want to do anything. I just want Codex to work on my behalf. That's it. So it has given us approximately exactly 12 names. So you can choose any of the name and give the prompt as use this name for my project. So let's see which name I like the most.
Face future mirror muse glow preview fine I'm going with face future it's good it is suiting our project idea so I will give the prompt as use the name as face future for my project here we have taken the idea from the codeex if you want you can give your name as well you don't need to take everything from codeex just I'm showing you you can automate all your task you can ask everything you want to know from codeex itself and codeex will perform it for you and we are pressing enter so it is working I also want to show you it's not always better to resume a session but you can go into a new session by pressing new if you want to start something new it is always highly recommended now I just want to show you how powerful codeex is by telling just one shot to develop a whole app for you.
So, ideally, this would probably be a mobile app, but I'm recording this on a computer. So, I'm going to show you a web app. So, for this, I'm going to give a prompt like, let's build this MVP as a web app. Ask me four concise questions to understand my vision for this based on what's missing in the readme or agents.md files. And I have given be concise because sometimes codeex or any other AI tools or agents exaggerate the things. So it is better to give be concise. So now press enter and it will build this MCP web app for us.
So it has asked me four questions as I've given him the instructions. So do you want real AI generated results now or a designed MVP with mock generated placeholders? So I want a real AI generated results for my app. Should the app feel more luxury beauty, clinical, professional or social viral? Something luxury beauty beauty. After upload, do you want users to see autogenerated 15 to 20 looks immediately or choose categories first? No, they should able to see immediately. You can give the prompt according to yourself how you want your app to be looked like daily.
And the last question is do you want the MVP to support only frontf face selfies or multiple angles too? I can say multiple angles and it will broader the scope of R. So it is better to give multiple angles. I wanted to use multiple angles. So I think I have uh given every answer to every question and it has enough information how I want my app to exactly look like. So let's press enter and let's see what it has to show us. It's cool features. This codeex tool by openai is seriously insane. And why I'm saying this you will see in few minutes when the app will be all ready and we can use it for our use or we can use it for making money once we will deploy it on the platform.
See if you want to speed the functioning of codeex then you can even press /fast. This will bring the codeex into fast mode and it will generate the results very fast. Here you are seeing that it is taking a bit time right. So if you want to generate results faster then use /f fast but currently it is working that's why I cannot give the command once it has done his working we can give the command it's not better not give in between see it is generating the entire code for us we don't have to write even a single line of code what used to take our few years ago is now being automated what you have to do is you can make the amendments or the changes in the app by changing few lines of code and you don't have to touch the entire code just that section if you want any changes or else you can keep as it is and you also do not have to manually change you just need to give the command to codeex and codeex will change according to your requirement.
It is super simple and super convenient just sitting and just giving right prompts and that is a real power of prompt engineering. So it is building some of the pages. Let's wait for this. And here it is shown how much time it has taken already. So as of now it has taken 2 minutes and 52 seconds. Let's wait for few minutes. It's worth waiting, right? Just wait for few more minutes. It is worth waiting until one of the amazing apps will be popped up on our screen. See the page flow is converted and the code is ready.
It is working on some more modules for the project and generating the codes in halves. So it is asking do you want to allow the production build to run outside the sandbox or a new multi agent Facebook? I'm giving yes approved. It is searching the web to generate the accurate most accurate results not accurate exactly. So the new flow is compiling and it is documented in next.js securely. See the another code is being generated. So another question it is asking do you want to allow installing the patch next.js version. Okay. Okay. I'll give yes and proceed.
So you are seeing that it is asking so many questions like I've already answered four five questions. It has asked it is good actually. So whatever the codeex is not doing anything by on its own. It is asking us how we want our app to work like. But sometimes it could be quite frustrating, right? That's why there's another cool feature of codeex is you can type codeex- yolo and then it will not ask any of the permissions. It will skip the permission part and generate us the code without asking any further questions. But it is highly not recommended because you want to see what is working on what all it is working the code is working right.
So it's better to let him ask how you want your app. Let you answer according to your requirement. Meanwhile you see how it is working in background and then the app will be ready. It would be more efficient to work like this. But if it is asking a lot of question, it is better to give the command like codeex - yolo and you are free. You can go, you can enjoy. Meanwhile, codeex will work on your behalf. Another question, do you want to allow dependency installation? Yes, I want. See, I have given the command /fast because I want it to perform a little faster.
But because it was working in background it has said it is disabled because something is going in the progress. So the another thing you need to note is you cannot give commands in between while it is working on something. So once it has completed its task then you can give any of the commands whether it is /fast / resume or codeex euro whatever you want to give and there are lot many commands you can type /help and you will find all the commands listed there it is very difficult to cover all the commands in just one tutorial so I would suggest you to sit for a while give it some time explore all the features perform it according to your project and then see how your projects is booming over the internet.
But the thing you need to do is you need to sit patiently and work with all the commands out there in codeex. And you can also take the reference of the official document of codeex which I'll try to attach the link in the description box below for you. But you can also visit directly by opening your web browser and typing the official documentation of codeex. So before working with any of the tools especially the new tools AI tools AI agent tools it is highly recommended that visit their documents first it it is such a amazing feature they have given like attaching the document with all the tools so that you have the basic idea about how it is working and if again you have any doubt you can visit to the YouTube channel of simply learn you can search over there and you must be finding any of the useful video on that topic.
So go explore on the codeex app and also the simply learns YouTube channel. It might take some time. As you can see it has taken approximately 15 minutes. So wait for a while because it is generating a huge and a complete f for us. So machines also take time. Whatever we use to code for 15 hours, it is taking just 15 or 16 minutes. So you already know how powerful is and how useful is to learn the code and the tools like this. See it is performing all the required task and installing all the dependencies and whatever is needed in this project.
So even if it founds any error, it identifies by itself and try to resolve it itself. No need of you. Just give a prompt and go enjoy. Come after 15 minutes or some minutes and then your project will be like half ready. Another question that is asking do you want to allow the production build to run outside the sandbox. Okay. So we have approved codeex to run npm run build this time. So would you like to run this following command? I want to press yes. Proceed. Here you can type codeex - yolo but it is working so we cannot give the command right now.
So now while this is working I want to explain IDE. So we will see how to use codeex with your IDE where IDE stands for integrated development environment and this is what you need when you are building any software. Now obviously for simple stuff for a simple demo or a simple project you can just stay in your terminal like this and here only you can do everything but if you want to build more advanced features or more more complex softwares or do you want to build more complex projects you want to use a proper ID you want to see the code if you want to see your file structure you can get so there then you can connect your ID with your codeex.
It is super fun and it is super simple and you just need to give you can use any ID whether it is cursor or VS code and you can give the prompt and it will perform the task for you. Meanwhile, it is asked another question. So I'm pressing enter. So where were we? Yeah. So I was explaining you you just you can just give the prompt and it will perform for you. And I will also show you practically how you can do it. And there are many benefits of using an ID and there is no reason to be scared of it.
It is just like any other app. So if you are not able to install like so if you're able to install Slack or Microsoft or Microsoft Teams or a video game on your computer, you can absolutely install any of the IDE. It is that simple. So you can even so the most popular ID in this world is Visual Studio Code. It is open source and completely free. But right now we also have cursor. It is more advanced way of working with IDE because it is AI integrated and has lot of AI features. So you can download any of the two IDs.
For my case I am using Visual Studio Code. You can even download cursor by visiting its official website. Just follow the steps shown on the desktop and in few minutes you will be good to go with cursor. Final build in next few minutes the app will be all ready I guess. So meanwhile it is performing. Go up see what all changes it has made. See the code see the dependencies and if you want to make any changes just skip the prompt and it will perform for you. See it has generated the entire code base the hundreds and thousands lines of code yolo because I want to see what all it is doing in the background.
So another question it has asked. So I'm giving yes proceed and it is loading much more dependencies. Whatever it is asking just read the question see the options like whatever you want to perform and give just I repeat and just hit enter. So it is asking the permission for running the following command. So run this. As of now it has taken 24 minutes to build the app. So yeah this is quite timet but very less as compared to what we used to do few years back. So it has edited the package JSON file for us.
It has performed the web search again to build in the accuracy for the project. Now I command is build the app for me fully professional and stop only when you are fully done. Now while it is working I want to just show you how you can open any of your project in IDE and work in the ID itself and not in the terminal. So here you have seen that we have done cd dot dot cd documents cdc codeex project. This is how we have navigated to our project file right. But here what we can do is we can directly open one of our ID and in my case I'm using Visual Studio Code.
So this is the interface of our Visual Studio Code. We can directly open the folder from here. Open folder our project folder. And here you can directly give the command as codeex and it will work inside your folder. You just need to select any of the folder open it in the Visual Studio Code. Then the terminal will show your project location. Here you can type codeex and you can literally build the complete app inside a proper folder integrated with your visual studio code. Start working with your ID because it is extremely good for complex projects.
You should build a habit to work inside ID and not in terminal. But right now I'm just showing a small project that is why I'm running it in the terminal itself. You can also run it in a terminal but try using IDE more. So again look at these hundreds of lines of code C codeex has generated yet it has taken a minimal time like 15 20 minutes but it has generated the entire code base for us without me having to intervene or guide it or anything. This is the power of codeex. It is the most persistent AI agent and can run for the longest and it can make the least mistake.
So I'm not joking when I say people who don't use codeex are missing out because it really is essential for building software. While it is generating the final project for us, I want to also give you information just to remind you please connect it with your GitHub account. Start a new git repository in this project and perform first commit. Do not ignore this. Do this. It will save your projects. It is performing the final cuts. Attaching the documents required. So here I'm giving the command as run this project for me in the default web browser.
I'm saying yes proceed. Just show me the project. I'm eagerly waiting for that. Boom. Our project is there on our screens. See the name of the project is face nature. Upload your angles. Get a fully beauty preview wall. Here it says zero accounts required. Four angles for your photograph is supported and 21 preset looks per session. So here you can upload one to four clear selfies. And you can see all the looks for your selfies from here. your smaller nose, filler lips, lifted outer eyes, raised brows, Botox, style smoothing, brighter under eyes and all the features.
You can even put these features into favorites list from here. So add your photos, see what all changes you want to have in your face. Select from here and before going to any surgery or any changes in your face, utilize this app for yourself. see this once whether it suits your face or not then take any decision. If you find it difficult, let's suppose you are not from the coding background, you don't know what all these codes are, then we can give the prompt to the codeex explain the complete flow of the project to me in a simple plain English.
Explain the complete project flow the plain English and it is working for us. So it has given us the complete functioning of our project step by step what all it has done what all we need to do and the quality status as well. If you want to make any updations in your app whatever we have seen on the screen you can give the amendments to codeex and codeex will generate it for you. And now once you are done with the entire setup, you will also have to connect it with your open AI API key which you will find in your shared GPT settings.
From there you can connect your project with your open API key. You can ask codeex to connect it for you. You can copy paste the API key. It will connect it to your project. So the only step that is left after this is connecting your project with your API key. And then you have to deploy your if you want to deploy your app and want to generate more money then you can ask give me the platforms where I can deploy my project it will give you all the platforms you can select according to your choice then you can ask codeex itself to deploy the project on any of the software and then you are good to go you don't need to do anything you just need to ask codeex to work on your behalf whatever you want to You can ask codeex to do it for you.
You ask the instructions how you can like deploy your app. It will give you all the instructions. It will deploy it for you by following those instructions by itself. Now I want to explain you about another command that is hyphen plan. So there are different modes in codeex. Now I want to go into a plan mode. It is one of the most important modes of codeex which will help you plan accordingly. See we don't want to generate the plan on ourselves. We just need to give our thoughts in a complete English like a prompt. Just put your thoughts here and it will generate the plan for you.
You don't need to generate the plan. For this you need to be in the plan mode first and by pressing hyphen plan as we have done before you can go and switch it into the plan mode. And if you want to see the changes you can run / review on your desktop. Now I also want you to shoot the desktop app of codeex. So I have already installed in my device. If you have not installed you can go to the official website of codeex. From there you can download the desktop version. So let me show it.
So from here you can go to the codeex app and you can download your app for Windows and then these are the steps you can follow. If you want to download it for Mac OS you can download it from here and for Linux then here then you need to open the codec sign in select your project and start functioning. So I have already downloaded it. So I'll show you. See this is how the codeex app looks like. I'll highly recommend you once you are done with the terminal CLI go here connect your project with this desktop app.
It is so so simple and userfriendly. It has amazing user interfaces. You can connect it with your GitHub, connect to your Gmail, connect to Slack and any of your favorite apps. You can type the prompt here. You can select your project from here. Whatever project you are working on, whatever you want to build, try something different, come up with some good ideas and design something really creative. You can check the plugins from here. And most importantly, you can go to the skills from here. You can select or download install any of the skills you want for your project.
Whether it is Sora, PDF, doc, image generation, open AI docs, whatever you want in your project, you can download all the skills from here. It is the newest version of how you can use codecs. Here we can also run different chats in parallel when if you are working on one project then you can work on another typing on new chat. So two chats will be working in simultaneously. If you want to work locally then you can tap on this and then work locally. It is my favorite way of using codeex. I want to tell you this.
And another thing I forgot to explain you is when you are going into skills and when you want to invoke that skill like let's suppose we have installed any of the skills you can use dollar sign shift dollar sign to invoke any of your skills. So this is the way you can invoke your skills. See skill creator skill installer and everything. So dollarer is the assigned to you skills. Here you also have automations. You can automate your work, summarize whatever you want. You can create different agents that will perform different different task for you simultaneously.
You don't have to wait or one agent will not affect your another task. So explore this app. Once you are done with the terminal interface, you can start working on this app as well. Now before we end let's summarize the most important best practices and the mistakes you should avoid in codeex learning journey. So the number one is always start with understanding before asking codeex to change code. Ask it to explain the project. Now number two use g checkpoints because codex can edit files. Git can help you safely review and compare the changes. Third is write specific prompts.
Do not say fix everything. Say exactly what is broken, what files matter, what should not be changed, and how to validate the fix. Number four, ask codeex to explain before editing. Number five, run tests. If tests do not exist, ask codeex to suggest test cases. And the last number six, avoid huge task at the beginning. Instead of saying build a full SAS F for me, start with smaller task. Start with creating the authentication module, then add the dashboard layout, then add building page, then add test. So these were some practices that you should follow and mistakes you should avoid in your codeex journey.
So let's recap what all we have covered in this video. Codex is OpenAI's AI coding agent that helps you write, review, debug, and ship code. It can work across your project files. It can understand your code base. It can make edits. It can run commands and it can help with real development workflows. You can use it through ID, CLI, web and app. The best way to use codeex is not by giving vague commands. The best way is to give clear task context, constraints, validation, steps and expected output. Use codeex to understand projects. Use it to build features.
Use it to debug issues. Use it to write test. Use it to review code. Use it to create documentation. But always remember this. Codeex can speed up your work. But you still need to think like a developer. Review your code, run the test, understand the changes, then ship the confidence and that is your complete beginner to practical course on OpenAI codeex. If you are a learning software developer, Codeex can help you understand code faster. If you are already a developer, Codeex can help you move faster across debugging, testing, documentation, and feature work. And if you work in a team, codeex can become part of your development workflow through code review, pull request, skills, and cloud tasks.
The future of coding is not just writing every line manually. It is knowing how to guide AI agents, review their work, and build better software faster. Now, the next step is very simple. Do not stop at just watching this tutorial. Open the tools, start experimenting, try small projects, break things, fix them and learn by building because that is where the real confidence comes from. If you enjoyed this course and found it helpful, do let us know in the comment section below. And if you want to build deeper skills in generative AI, then do check out the program link in the description box and in the pin comments.
Thanks a lot for watching. Keep learning, keep building, and I will see you in the next one. Thank you.
More from Simplilearn
Get daily recaps from
Simplilearn
AI-powered summaries delivered to your inbox. Save hours every week while staying fully informed.

![Power BI With AI Full Course 2026 [FREE] | Power BI With AI Tutorial | Power BI Course | Simplilearn thumbnail](https://rewiz.app/images?url=https://i.ytimg.com/vi/rsPa0xeLrYY/maxresdefault.jpg)
![Power BI With AI Full Course 2026 [FREE] | Power BI With AI Tutorial | Power BI Course | Simplilearn thumbnail](https://rewiz.app/images?url=https://i.ytimg.com/vi/RUpsX299R4k/maxresdefault.jpg)






