Free Claude Code + Ominroute is INSANE!

Today I'm going to show you how to use Claude Code for free, forever, with a new setup that's just dropped from OmniRoot. OmniRoot basically allows you to code with over 231 free providers with one single endpoint. And then you can plug that into something like Free Claude Code or into Codex, and you can code for free, forever, with this agent harness. So, for example, if we go inside here, this is our agent OS, and we've already plugged Free Claude Code into it.

And Free Claude Code is an open source project that you can point to any brain. So we've plugged it and pointed it at the OmniRoot brain. So if we go inside here and we're like, OK, build out an SEO agency website, like so, it will start to build out, as you can imagine. And then if we want to see what we've built, so, for example, we can wait for that to start working over here.

And then if we want to see what we've built previously, we can just go into Free Claude Code, go into the workspace over here, and we can see all the stuff that we've built. And it can build some pretty nice stuff. Like, for example, this website we created with Free Claude Code and OmniRoot. Here's another example.

Here's another example of an app we've built. And this is pretty easy and simple to do. In fact, we've got loads of stuff that we've built out here with Free Claude Code. Now, out of everything that I've tested, it's pretty interesting because out of everything that I've tested, and I've tried many different free APIs inside Free Claude Code and also local models, and I would say OmniRoot by far has been the nicest to create stuff.

So whilst that's working in the background, let me show you another example. We actually plugged this into Codex over here, and we can ask Codex to work in the same way. It's using OmniRoot. And then if we go to our workspace here, we can see everything that we've built for free with OmniRoot as well, using Codex, right?

And it looks super nice. Some of this stuff is really nice. Now, how does this work? Well, basically, this is the open source project, the GitHub, and essentially it means that you can use the coder directly.

It's a free AI gateway. It connects to, like, 237 providers. Ninety of them are free, and it works through one endpoint. Now, you can plug Claude Code or Codex or Cursor or Client, Copilot, even Antigravity into Free Claude Code.

The great thing about that is, like, a lot of people are worried about rate limits on free APIs or, for example, using local, and sometimes local is too slow or it's not powerful enough. Or, for example, you don't get the sort of frontier outputs that you want from something you're normally used to. The other cool thing about this is when it's being used, it combines RTK and Caveman, which are two ways to compress the tokens you use to save, like, 15 to 95% tokens. And additionally, when you're coding out with this stuff, the other great thing about it is that it automatically falls back.

So, for example, if one API stops working or isn't available or is rate limited, no problem, it can switch to the next one over and over again. And so this is a pretty powerful way to use Claude Code for free with this setup. And so I've called it the Free Claude Code Engine. And so you're using, like, the actual Claude Code CLI, but it's rooted through OmniRoot across 93 providers.

And you will also fall back when one runs dry. Also, you don't need to subscribe to Claude to use this. And you can embed it inside your favorite operating system. So, for me, for example, we have the agent operating system of Claude, OpenClaude, Hermes, everything else.

And then we have a separate section for Free Claude Code, for Codex, and also for OmniRoot. Now, you can also code directly with OmniRoot here, as you can see. And what this allows you to do is you can use the same system, but you can basically see a preview of what you've built. Then you've got the saved stuff that you've created.

So, for example, here, we've got all the mini apps that we've built. One thing I will say is you're not going to get, like, Fable 5 level outputs with this. But if you just want something to code out some quick ideas, or build some mini apps, or, for example, create, like, a few landing pages and blogs, et cetera, that's perfect for this. Like, that's actually going to work quite nicely.

And the great thing about this is you don't hit limits. You save 95% of your tokens. It's $0 to start. Every tool actually works.

It works with over 24 coding agents, which is pretty wild. There's one endpoint, and it's production-grade as well. So, like before, a lot of people love to Claude code. But every session is limited by the number of tokens you have available.

And quite often, particularly if you're using Claude for a lot of big projects, you just run out halfway through, or you get token limited. So, for example, we go directly to Claude over here. You can see it says approaching session usage limit. So we are limited in terms of what we can build, because even though we're subscribed, we have a limited number of tokens.

With this setup, we can use the exact same Claude CLI, but it runs through 93 providers. And if one runs dry, it hops over to the next, so we never get cut off. We can build all day, and we don't need to worry about the token limits. And it's a tab inside the AgentOS.

So if I type, it writes files, and I can preview them live. So you can see, for example, it's actually built out the full-size structure for what we created. And then if we actually want to see what it's building so far, we can see that over here. So you can see when this was updated, like two minutes ago, two minutes ago.

We've got the outputs inside the chat, and it's just building and working in the background here. And then over here, we can actually preview what it's done so far. So for example, if we click on this page, it's already built out the services page, which looks super nice, and it's easy to use. Then we can see the source.

We can open that up in a new tab if we want to as well. And this is all coded out locally. So what that means is, like, you're using free APIs in the cloud, but the actual outputs, the files that you build, the HTML, the code, et cetera, is coded locally, so you can come back to it whenever you want inside your workspace. So if you're like, oh, where did that page go?

Well, you can see it now. And you can see that it's actually building out the full blog as well, which is pretty cool. So it's building out, like, a full website, organized it, actually looks quite good. We can go back and forth inside the chat as well if we want to.

And the other benefit of this is, like, because it's linked to our memory setup, so it's using our full Obsidian Memory Knowledge Center here, it can use that to create better outputs inside the chat here, right? So if you have ideas on, like, how you want the voice to be or how you want the page to look, you can actually add style guidelines inside your memory, and then free cloud code can link to that and use that to actually write better outputs. So that's pretty powerful, pretty easy stuff. And then the way this works, it works in five layers.

So you've got Cloud Code CLI. That's running through OmniRoom. You get 93 providers, and then you get your build. And that all works inside one tab inside the HNOS.

So you get the prompt section, a workspace, and a live preview. Now, you also might say, okay, well, three models don't match the real cloud. What I would say with that is, like, again, you're not going to get Fable 5 level outputs with this, but no other model matches that anyway. But what you will get is something you can actually use that's actually useful, that actually looks good.

You can build out full projects, and it's easy to set up as well. So it's literally coded this whole website whilst I'm talking to you. And if you look at the page, like, it looks pretty good. And you can always improve this as well.

So you can go back and forth inside the chat. You can tell it what you want to improve, what you want to change, et cetera. But it's all set up and ready to go. And that's a full website, fully automated as one project with free cloud code.

So free models can still create really good stuff. And the other difference here is, like, when you're using this, you don't need frontier models for everything. So if you're just coding out a simple landing page or a website or a blog post, you don't need frontier models for that. You can just use free APIs like this.

Now, also, there's a bunch of settings that you can change to. So, for example, with OmniRoot, you can optimize depending on what you're doing. Like, if you just need fast outputs, you can change to auto forward slash fast as the model ID. If you want really good coding outputs, then you can switch to auto forward slash coding, and it focuses on quality coding generators.

So depending on what you want, you can switch. You can even switch to offline as well, which is pretty cool. And then you've got a bunch of different routing strategies. So depending on how you want the models to be used and which models you want, you can set that up as well.

And so if you look at the old way, you would be using a lot more tokens, and you'd be worried about everything you're creating. With the new way, it's like, I can build out a full website. I don't need to worry because it's on free APIs anyway. And if I do need to change anything, we can do that later.

You can also see here, for example, that you can run free Cloud Code in about five minutes. So how you do this is use this one command. Then you're going to switch it to Cloud Code, and then it's just going to be one tab inside our Agent OS. Now, some people say, you know, I'm not technical.

This sounds difficult to set up. So what you can actually do is you can just use these two terminal commands, pretty simple and easy. Or the other option that you have is that you can actually go into Cloud, and you can give it the details of the GitHub and ask it to set that up for you with Cloud Code. So if you're not sure how to configure it, what I did is just say, right, here's a GitHub.

Set this up in Cloud Code inside our Agent OS. And it actually went off and did it, and it was super smooth. So that's basically how it works. And then Cloud Code will route to multiple different free providers.

You don't need to be technical to set it up. And also, if you're using an Agent OS, and you've got loads of CLIs in there, you can build loads of cool stuff. But you don't need to worry about tokens, because you're using free APIs. And then for like the more frontiers level stuff, of course, then you can switch to normal Cloud Code.

So thanks so much for watching. If you are wondering, like, is this technical to set up, et cetera, we've got so many people getting amazing results with the Agent OS. As you can see, we actually got over 196 pages of people winning, learning, and growing with our Agent operating system. So if you want to get that, that's inside the airprofit warning link in the comments And inside there, you get the full Agent operating system that you can see.

So you get free Cloud Code, the workspace. You get OmniRoot that we've plugged in right there. We also have, for example, Codex plugged in. We've got Cloud ready to go whenever we need it.

We could also use this inside Hermes as well. And then we've got Paperclip to orchestrate all of our agents, a group chat with agents, a pipeline for going from idea to implementation really quickly. And then also inside Hermes, we have the Oracle. We have Astros for finding the latest trending topics.

And we also have Apollo, which is a voice-activated version of Hermes as well. So if you want to get all of that, it's inside the airprofit warning link in the comments description. Or just go to the airprofitwarning.com. Inside the community, you can ask questions, get help and support in real time.

I personally answer every single question and answer it with a new video tutorial each day. And then inside the classroom, you get access to all of our best training. So if you're a complete beginner, you can go from beginner to expert over here. If you want to get our new daily updates, you can get that over here.

And we also have the full Agent OS system. You can get a video tutorial for that and see when it was last updated. You get the zip file. And we also add new daily tutorials based on what's actually working.

So we have a full tutorial on how to use OmniRoot and how to implement it step by step. And every time something new and useful comes out like this, I test it, and then I'll show you how to use it. Inside the calendar, you can jump a weekly coaching course, get help and support in real time. Inside the map, you can meet people in your local area who are building with AI agents like you.

And that's all available inside the Air Profit Boardroom. Link in the comments description. Or just go to the airprofitboardroom.com.