Run GPT-6 Astra + Claude Fable 5.1 On This Free AI Agent Workspace transcript

Stop chasing every single new AI tool because this free AI agent workspace allows you to run every single model on your work without having to restart from scratch. Annoyed with Cloud Opus 5 or suddenly there's new GPT Astra 6 that's popping up? Not to worry because you can now swap these models in very quickly and your business only improves with the new models. So in this video I'm going to show you guys how you can do it and how you can set this up all for yourself.

So without further ado, let's get right into it. So this AI agent workspace is called Nami and I've created this over the weekend because I wanted a nicer space to run all of my different agents work and be able to swap in the different models as and when I like. This is how it actually looks like inside my computer. You can see I have different sessions from different models whether is it Grok, Q me, Codex or Claude and I can use it to run any of my work within my workspace itself here.

These are my AI OS which contains all the different contents, my thumbnails and every single work files that the agent can use to generate work for me. If there's one thing I learned over the past three years of being inside AI is that you don't want to be tied down to one single model. You want your business to be able to move whenever a new model comes up. And so I built my whole AI OS around things that really matters so that any single model can actually access it but I still needed a space to be able to use the different models to visualize my work nicely for me.

I spun this up over the weekend so that I can use it for myself and you guys can try this out as well. You guys can just get it over here. It's completely free and open source and then you can just open it up on your own computer. You can open up a workspace for agents to be working inside of it. So I'm going to open up a folder here and I'm going to choose my folder Dynami OS which is my AI operating system folder and open this up.

Okay. Once I have that all of my files are on the left hand side and I can view all of them over here. For example, this are all my content folders. You can see all my different folders for the different contents like short form, LinkedIn, YouTube and so much more. Okay. Once I have this, I can point any single agent at my folder to work for me. And to do so it's super simple. I just have to click a new session and I can choose one of the different models that I want to use.

For example, Codex, Claude and run them against my folder. If I want to set up a new Claude code session, I just have to click into this and here I have a new Claude code session. Right, I can set up a Codex session beside it as well. Here we have it, Astra 6, and then we can also open up, let's say, Grok, or even if we want to do something like Kimi, we can have all four running our work all together side by side. And I can see all the sessions on the left-hand side over here under sessions, and I can see what's going on with the different agents itself.

And it's super cool because I'm not exactly tied to one single model. Now, if I want to use Codex for a specific task and Claude for another, I could just do so by creating two different sessions side by side. And because all my work are inside my AI OS on the left-hand side, any of these agents will be able to pick them up very quickly and do the work exactly in my own brand tone and following my own set of instructions.

Okay, now that we have all four sessions set up, it means that we can make all four of them do the exact same task for us within our folder. Inside my own folder, I have this thumbnail, okay? And I want to do a new thumbnail job and I want to get every single one of these agents to do the exact same thing for me. Because I already set up the skills and MCPs correctly within my folder, all the different models will be able to take on the job.

So, I'm going to copy this link here, okay? I'm going to copy this link here. I'm going to paste this in and I'm going to say, "Hey, can you create a new version of this thumbnail using the ChatGPT image to replace the Claude image instead. Make sure to use the correct assets to generate this thumbnail." Okay, click enter. And I'm going to copy this exact same command and I'm going to paste this in into every single other section.

Paste this in. Paste this in. Paste this in. All right. And now all four models will be working for me to generate this new thumbnail. So, let's give it a moment and see what we get out of this. Okay, and it's done. All of them have returned the thumbnail for us. So, let's take a look. This one is by Kimi. And here we have the ChatGPT logo. Of course, it uses the same colors, so we might have to tell it once again. But yeah, we have the job done by Kimi.

We also got Grok to do the same task for us because he is able to call the MCP as well based of our skill. And here we have our same face, right with a black version of the chat GPT logo instead. And we also have one from Codex. Here we have it, the white logo, yeah. And another one from Cloud Code, which is pretty similar to the one from Codex as well. But hopefully it showcase to you guys how I'm able to run four different sessions over here with four different models to do the same exact task for me for my own work.

And that's all possible because I built my system around the things that matters, the skills, the instructions. And I'm able to point every single one of these models inside my workspace very easily towards my folder. Besides that, I want this workspace to be simple for myself to run all my different agents as well as being able to view it in the way I like. So I can always expand one of these sessions up here just like that, you know, and then go through each of them.

I could open up any of my files here, right? And it has this nice view. I can pin it to desk as well. And then I can look at my image or my files while having my agents run it for me. Okay, and all of them will be here in the sessions tab. And lastly, one of the things that really prompted me to design this over the weekend is that I wanted to be able to view all of these different sessions in a way that's nice and fun.

The terminal view doesn't have to be scary, so I created a bunch of themes as well that you guys can try. For example, you guys might want to take a look at terminal in a paper view. You guys can try this out as well. And you guys can change it up to other views like the glass view or you can change it up to, let's say, a graphite view like this, right? And other views that you like like the operator view that I like very much.

If you guys want to run different agents like how I'm doing on your own workspace, you guys can try out Nami and let me know what you guys think. And before we end off this video, if there's one thing I hope you guys can take away from this is you want to make sure your business is always built to be model agnostic. And it's not as difficult as it seems if you focus on the things that truly matters like the skills, the instructions, and so much more.

If you guys want to learn how I actually built this up, I have a full guide on how to build your own AI OS, you can check it out over here. If not, with that being said, I hope you guys enjoyed this video. If you like videos like this, do give it a like and subscribe, and I'll see you guys in the next one. Peace.