Hermes Mixture of Agents is ABSURD!

Hermes mixture of agents is absolutely wild. So this is a new system where you can combine different models. And the goal here is that you have a panel of agents that work together and get better levels of intelligence than frontier models. So you can see some examples of what we created here.

So for example, we actually created a full windows style operating system. Looks absolutely beautiful. And this was created using mixture of agents from Hermes. Now, Hermes itself is just an agent.

But what we're doing here is we're mixing different models together. So for example, we have Claude Obers 4.8 and we have GPT 5.5. And then we can choose who is the aggregator. Now, if you're wondering what is an aggregator, that is basically the model that will fuse the answers together.

And then from here, we've created this system where we can basically ask the panel anything. So if we type in, for example, a prompt here, we can then run the panel and then work together. And whatever we create goes directly into our workspace. And it's creating some pretty amazing stuff, as you can see right here.

Now, if you're wondering how does this work? Well, the main thing that I would say here is that the model doesn't matter anymore, but the system does. And this research, who actually created Hermes agent, basically now allows you to stack several models inside Hermes. And this can beat Opus and GPT, not with a bigger or better model, but with a smarter system.

So let me explain how this works step by step. And by the way, if you're wondering, OK, like, have I tested this or do you have a demo, etc.? Do you have side-by-side comparisons? So we actually tested 42 different builds with Hermes mixture of agents.

And again, this is a panel of Frontier models merged by a chair, which means that you have a system that works together. Now, with this system, we've created loads of different builds. You can actually see it on GoldieBench. And then what we can actually do is compare this side-by-side versus something like Claude Opus 4.8.

So this itself was built with this mixture of agent system. So the point here is like Opus 4.8, amazing model, fantastic, probably the best one out there. But when you have two models working together, they're always going to outperform one model working alone because two minds are greater than one. So if we compare Opus 4.8 versus Hermes mixture of agents, there's actually quite a big difference here.

So let's take a look at these two examples. So this is a crypt game that we created. Looks super nice, great background, etc. And this was created with mixture of agents.

Now, we actually asked Opus to do the same thing, and it failed. We can't get through the start screen, as you can see. Now, it didn't fail on everything. So if we compare these side-by-side, this is created with mixture of agents.

Again, looks super cool. Kind of feels like the original game, etc. Actually a lot of fun to play. And then if we compare that side-by-side versus Opus 4.8, it doesn't feel quite as nice.

I mean, it's still good. But that's the point, is if you want the best, then you would combine multiple agents. And the point to note here is like, mixture of agents is not a model, it's a system. Whereas Opus 4.8, that model will change.

And the thing here to note as well is that whatever models change and whatever models come out next, so for example, like Sonic 5 is rumored to come out next, then we can combine that with GPT-5 or GPT-5.6 when it comes out and get even better outputs. So the whole point is that we get higher levels and better levels of intelligence using the system. Now, when you're running it, as you can see here, the panel will deliberate, and then you just have to wait for the outputs. The thing that I would say here as well is it's not all perfect, because when you're using mixture of agents, it can take quite a bit longer to number one, get both answers from both models and then fuse them together to get a good answer.

So it can take a little while. And also these would both run on APIs, of course, as well. So Opus 4.8, you can't use with the CLI if you're using mixture of agents. So those are two honest disadvantages to just make sure you understand before we go further.

But the point is, if you can get better levels of intelligence, well, that's absolutely awesome. Now, the other thing to note here is like if you're using mixture of agents normally, for example, inside the terminal, it's very difficult to manage them. And it's quite technical. Let me show you an example of the instructions for using this.

So this is what you would normally do. You'd run loads of terminal commands, and you can select between CLI and the gateway. Then you run more terminal commands. Then you follow like seven different steps.

Then you do a lot of coding. So it's not that accessible. And that's why we actually built it as a tool inside the agent operating system that we have here. And I think that's a much smarter, better way to use it, because you can see everything that you've created.

You can preview it. You can save it for later. And the great thing about that is you don't need to be technical to use it. It makes something technical very simple for almost anyone to use.

And it looks absolutely awesome. So what just happened here? Well, Noose Research dropped mixture agents of 2.0 inside Hermes agent. So you can pick a few models, any providers mixed together.

That could be open. It's GPT 5.5, an open model, whatever. You save that mix as a preset, and then Hermes treats a whole mix as one model. So under the hood, the other models do private analysis first.

Then one aggregator model reads all of it and writes a final answer and runs your tools. So you flip between mixes mid chat with forward slash MOA, and it keeps prompt caching so it stays fast. And the result Noose is claiming is that it's 8% higher than Opus 4.8 and 11% higher than GPT 5.5 on benchmarks. So it's not a new model.

It's just a new system built on the models we already have. And we saw this in a similar way with Fusion, which also gets great output. So if we go to Fusion, this works in a similar way, where basically you can have a panel of agents, a panel of models working together, and they tend to create something better than if you, for example, had just one model working alone. So this is a proven system that some people have been using for quite a while now, but it's really something that's becoming easier and easier to use as Fusion gets released, as Hermes releases a mixture of agents, and also as more and more frontier models get gated.

So for example, everyone's waiting for Fable 5 to come back, but in the meantime, they don't realize actually you can use a system like this and get better levels of intelligence, better outputs, without having to wait for that model to come out. And that was really their point. You know, the strongest models are gated. This is a tweet by News Research, and access is granted to only a select few.

So Hermes Agent now exposes a mixture of agent presets as virtual models, giving you capabilities beyond the publicly available frontier. So if you want better than the frontier right now, you can use this system. And so you just give it one prompt. You get three different models working together.

You get one aggregator. You can just have two models working together as well if you prefer. And then you stop chasing the models, you build the system. And everyone is waiting on the next model, the next Opus, the next GPT, the thing that will finally change everything.

But if you look at these models getting gated, that might not be an option in the future. And so this is a better way to combine the models that you do have, especially open source models and cheaper models to get better outputs. The other advantage of this as well, of course, is that you can use cheaper models working together and potentially get better outputs than if you had an expensive API, which is pretty mind-blowing in itself. Because then you can get like super high level outputs, but you don't need to spend money on super expensive models.

And so there's three different systems that have come out recently using this mixture of agents from Hermes, Fusion, and Sakana Fugu. And they all use the same models, the same systems, the same panel engine. And that's what's getting them better outputs. I mean, for example, Fusion actually outperforms Claude Fable 5 on benchmarks.

Now test this stuff for yourself. See what you think. You know, you don't have to listen to benchmarks, but the point here is it's a really good option, especially during times like this. And so if you look at the old ways, like you would wait for the new model to come out, hit that ceiling, get rate limited, wait for access to frontier models, pay top dollar for the most expensive APIs.

But now with mixture of agents, you can basically mix several models, even cheaper models, beat the best single model, get frontier level intelligence without having frontier models, go beyond the public frontier without extra access needed. You can start using it today. And again, like you might say, well, this sounds complicated to run, but actually if you have a system like this and we give this away to AR Profit Board members, link in the comments description, or go to the arprofitboard.com, then you can use this in a simplified tool where you just ask the panel anything and it's ready to go whenever you need it. So that's basically it for mixed use of agents.

Now, if you want Hermes agent with a panel built in, and we've also got mixture of agents, Fusion, Sakana, all running on the same idea. They all live inside our agent operating system, which is a dashboard I can run my whole business on. So you get the full agent operating zip, Hermes, mixture of agents, Fusion, Sakana. We've even got, for example, GLM 5.2.

We've got code over here and we've got a memory galaxy so that all your agents and whatever you use has context from you. Plus we've created tons of great custom automations directly for Hermes, as you can see, including the lead generation tool, Hermes Oracle system for news and Hermes Jarvis, which is a voice activated version of Jarvis and Hermes. Then you can get that inside the AR Profit Board room. Link in the comments description or go to the arprofitboard.com.

And inside this community, you can get access to an amazing community of people building amazing stuff with it. So I mean, for example, we actually had Robbie who posted today about using the AgentOS to pitch for freelance jobs and to land freelance jobs. So it's a fantastic community. You can ask questions and I personally answer them with a video tutorial every single day.

Plus we always have people online so you can get help and support whenever you need it. Inside the classroom, you get access to all my best trainings about AI automation. And if you want to get the AgentOS system, you can get it over here, as you can see. And then we add new daily tutorials like you can see right here.

We also update the AgentOS every single day. So you get the latest update, you can get the zip file to install it and a video tutorial plus a guide on how it works. And you can also get four weekly coaching calls where you can get help and support in real time. And you can meet people in your local area from the map so that you can connect with people near you who are building similar things.

So thanks for watching. That is a mixture of agents from Hermes. See you on the next one. Cheers, bye-bye.