Today we have a brand new update from Hermes as they've released Hermes Agent V0.13. So let's get straight into it. This is the Tenacity release and
we'll walk through exactly why that is in a second. If you want to update your Hermes agent, just go to Hermes update. So for example, we can go over
to our terminal here and we can type Hermes update and then we can update to the latest version and
get that running in the background. In the meantime, let's have a look at what's changed, what's new, why this is, etc. So, this is the Tenacity
release. So, Hermes agent now finishes what it starts, right? And this is something called uh a multi- aent cam board inside Hermes, right? So, if
you've not come across that before, I'll show you exactly how that works in a second. There's also forward
slashgo, which keeps your agents um focused on a turnbased release style. Basically, that's like a Ralph loop if you ever come across Ralph loops that
dropped at the start of this year. And it's a really powerful way to use AI agents. So, let's look at the highlights and also look at this agent
update in here. So, we've got the new release as you can see and it's just restarted the gateway. So, if
we type in Hermes now, we can start running this with Hermes agent and you can see we have updated v0.13. So, just to recap, to update, you type in
Hermes update inside your terminal if you already have it installed. If you don't have it installed, you can just install it directly from GitHub. All
right, so let's talk about the highlights and what we've got here. So, basically, Hermes, they
recently released something called Kamb, right? And if you're not familiar with this, basically you can go to Hermes dashboard, which is the web UI
for Hermes. And if we go over to Kambban here, this is a Kambban board, right? And so with a Kambban board, you can move tasks across different
sections like you can see here, right? And so for example, we recently created a new AI SEO blog. We built a
new AI SEO content strategy. We built out a new website using this camb feature. Now, what you can actually do here is you can have multicambans. So
for example, you can switch between assignees here. You can switch between boards and you can set up new boards. So what does that mean? Well,
basically you have multi- aent profile collaboration. So you have different boards and multi- aent camb
boards if that makes sense. So you can have multiple projects on the same board or you can have multiple boards with multiple projects. I don't know
if that makes sense. If you're watching this live, just feel free to ask any questions as we go along so that I cover all the gaps and also I help you
as much as I can. But essentially that means that you can have different Hermes agents like
different profiles and you can have different boards and you can have different projects moving across the board in different tasks. And so what this
basically means is like you're not just having one AI agent, you've got a huge team of autonomous agents all completing multiple tasks at the same
time. So you don't have to wait for Hermes to do one task and then move on to the next. Right? And so
that's more autonomous and more powerful than maybe anything we've ever seen before, if that makes sense. Now let's come on to some of the other improvements and what's happened here.
All right. So, we've covered multi- aent cambban. Basically, you can delegate to an AI team that actually does stuff. If you're wondering, okay, what
does that look like? Here's an example. So, we said go off and create this camban board. Like, you can see um we said build a an AI SEO website as
well. And then you can actually see the logs of what it's done. And then inside the chat as well, it
will tell you what it's done, right? So, for example, here actually built out a full 50page website for us and deployed it. So, this is the website
that we deployed. I actually got a separate video on cambbans with Hermes agent, but basically we built out this full website using the camban board
and Hermes agents. So, it's super powerful. Now, you've also got forward slashgo. And what this means
essentially is you have an autonomous agent looping over until it completes a task, right? So, let me show you some examples of this. It's going to
lock on onto a goal and use at least 20 turns until it's completed using this new update. So, let me show you how that works in practice. So you can
go over to Hermes like so and you can type forward slashgoal and then you give it a goal to work
towards. So let's say for example we were like okay I need to build out an SEO blog for my website. create five pages or create a beautiful design for
the blog first in HTML so I can preview it. It's about AI automation, make it look beautiful, etc. Right. So, we've used the forward/goal feature
there and now the goal has been set. Perfect.
So Mister says, "I feel like I'm in the stone age using Telegram and group topics. Hermes updates are moving faster than I can react." Yeah, there are
a lot of new updates. One thing I would say is like actually it's easier to keep up with Hermes than it is with OpenClaw. And with OpenClaw, if you're
using that as an AI agent, that's more it breaks a lot more because they're updating and releasing
new versions every single day. If you're using something like Hermes and V0.13, the cool thing about that is that when it's being used, it's only a
weekly update. So, there's a lot of updates inside that weekly update, but it's only once a week, which is good. that saves a lot of time and a lot of
focus as well. So basically you can give it this goal and what that's going to do is focus on
completing 20 loops to complete that. All right. So we've talked about the multi- aent cambban and how you can build multiple boards with multiple
agents working towards multiple tasks. I've shown you an example of what that looks like in practice. For example, the 50page website that we built.
Now let's talk about the next thing um in a second but first as well we just need to talk about forward
slashgoal and this means the agent works towards its task. So that's two highlights right there that we've already covered here right and we've got it
working towards the goal as you can see now also it's got a new video analyze tool and basically what this means is it can analyze videos. Now, if you
weren't familiar with this already, they also have another video update which means that it can
use hyperframes, right? So, it can number one understand videos which is awesome inside Hermes agent and also the other cool thing about this is that
it can generate videos as well. It can create videos inside Hermes agent. So, if you wanted to see an example of that, let me show you how that works.
So we can go into Hermes on another terminal here and we can say okay create a I we'll just create
a super short super short amazing animation about Hermes agent using hyperframes skill right and so it uses this skill to generate awesome videos if
you've not seen that before that is another new feature that came out all right so there's there's so much cool stuff the other thing I would say here
is like it actually works. That's what we like about this, right? Like when you do this, it actually
works and actually does stuff. Whereas, for example, if you're using uh something like open claw, quite often it doesn't work, you know, or it does
break. Um so says it doesn't break if you don't update. That there's the key there. There's the key. I've said that on a few videos recently with
OpenClaw. It is more reliable if you don't update it as often, right? In fact, sometimes it's better not
to update these tools at all if you're happy with your existing setup and you don't want to break anything. So, little tip for you there. All right.
So, we've covered multi- aent camb, we've covered goals, and we've covered the video. Now, you can also use X AI custom voices as a texttospech
provider, and this can clone voices as well, right? Which is pretty powerful, too. So, that's a whole new
feature that's been added. That is feature number four on the list. We've also got new language updates. So, it has a translation option for Chinese,
Japanese, German, Spanish, French, and all sorts of other languages, too. So, there's seven new languages inside Hermes as well. We've got Google
Chat. That is the 20th messaging platform. So you can actually speak with Hermes inside Google chat. By
the way, you can see it working on the goal here and what it's doing. So it's actually building out this task right here whilst I'm talking to you and
it's working towards that goal, which is pretty cool. Let's see where we're up to on this. Yeah, so it's building out a markdown file for the Hermes
um design of the website, which is pretty cool. And it's building out the video as you can see right
here. So we have two things going on in the background here. We have the goal and this is one of the other cool things about Hermes. Like you can have
multiple Hermes terminals working at the same time completing completely different tasks. So as an example that we've got two different things here.
We've got the video being created in this terminal window. And inside this terminal window we have
the website being created using the goal feature. So if we go over here and we go to this page, let's open it up. Boom. Look at that. It's just built
out the website um and hosted it locally as you can see right now. I would from there I would actually give it the feedback that you need to improve
this and make it look better. But you can see how we've quickly created this page right here. I mean
it looks all right. It's not bad but it could be better. It could be better my friend. Um so yeah our whole website using the goal feature. That's an
example of how you can use it. So you can see how powerful it is. Obviously, you want to look at what are you spending your time on and then give
Hermes agents the tasks that you spend the most time on so that you can automate those, right? And the
difference between using Hermes agent to do this and using other tools to do this is that Hermes will just keep going at that task for like 20 times
if you want. And then it actually has a judge that checks if it's complete or not. So it actually says at the bottom here, goal achieved. The response
confirms the HTML blog design preview was created. So actually created the whole website using this
process. Pretty amazing right there. My current workflow is to have Hermes update OpenClaw. That's a great example of how you can have Hermes working
directly with OpenClaw. It doesn't stop until it is working. That's a really good way to do it. I think that's a very smart way. So yeah, good for
you, man. All right, so we've covered cambans, goals, video generation, voice and text to speech
updates, the new languages, Google chat. There's also a bunch of like what I would say is like the more boring stuff. So you've got like um sessions
that survive restarts. So if you had a session with Hermes and then you restart it, it will remember the previous conversation and auto resume, right?
Which is pretty cool. You've also got a bunch of security fixes. You've got checkpoints as well. It
links its own rights. So let's have a look at that. So basically, I think it just it improves the way it writes its files, which is better as well.
Then we have improvements on scheduled tasks. more allowment lifts and also you can use Hermes as a MCP right so if you have a lot of tools connected
inside Hermes you can give that MCP access to any other tools for example like it could be claw
desktop it could be for example something like uh codeex right you can use all these other platforms to control Hermes agent using the MCP server from
Hermes agent and Hermes MCP is just like connection, right? It's just a way of connecting one tool to another. So, if you have loads of tools and
loads of powerful skills inside Hermes and you're like, I don't want to install that all over again to
codeex or something like that, you just give the MCP access from Hermes to whatever tool you want to connect it to and then they can communicate
together and it becomes way more powerful. Also got TUI. So, model picker matches uh Hermes model. The dashboard has improved as well. So you can now
have a there's a new plugins page inside there as well. Let's have a look at that as well. So if we go to
the dashboard inside Hermes, which is here, you can run Hermes dashboard as a command to check this out. There should be a new plugins feature over
here. Let's find it. Yeah, there we go. So you've got plugins here. You can manage your memory provider. As you can see, we've got Honcho set up
context engine. So like we've set up compressor right there. You can install from a GitHub or Git URL if
you want to. And you can manage all of your plugins down here, right? So, you see how it's got like this Google Meet plugin. Uh Spotify was on the
last week's update, which is pretty cool. All sorts of extensions you can install right there. So, you can manage your plugins from one place. Plugins
are just like ways that you can connect up tools like you can see right here. Right. Pretty nice. But
when are they going to release chat inside here? That's what I want to know because right now I'm using Hermes Workspace and Hermes Workspace is good,
but I feel like why haven't Hermes agent released her own sort of UI for chat inside here? There must be a reason, but I'm not sure what it is. Um, so
we got multi. Um, yeah, this is interesting as well. You can share board workspaces and work logs
across profiles as well. So if you got one agent working on your camban, well then that camban board and that task and that history of the task and
the context of it can be shared across multiple Hermes profiles. So you have multiple different Hermes agents working on the camb and understanding
what's going on. The goal section we've talked about improvements, messaging, etc. I think that is
pretty much it. Let's have a look here.
Yeah, I think that's pretty much it. So, let's let's recap on everything. What is Hermes Agent? Oh, it's just an agent that runs autonomously locally.
It's free. It's open source. It's an AI tool built by Noose Research. Open source means anyone in the world can download it and use it for free
basically. And you can give it instructions and autonomously goes off and does stuff. Now, Hermes got a
huge upgrade, which is version 0.13, nicknamed the Tenacity release. Tenacity means never giving up. And that's exactly what this update is about,
Brent. That's what we're talking about here. Now, why should you care about this update specifically? Because if you run a business, you need things
done fast. Right now, hiring a team requires a lot of time, a lot of energy, a lot of resources. But you
can have AI as a team now, and you can have AI working together as a team for free now. And that's what Hermes Agent does because the new version
makes it smarter, makes it faster, makes it way more reliable than before, right? They also closed 282 issues and improved it 588 times in just one
week using this update. It's a pretty big update. And here's the 10 things that have changed. Right? It's
a multi- aent cambban. So you got a cambban board like a to-do list on a whiteboard with columns like to-do, doing, and done. Right? And now Hermes
agent can run multiple AI workers on that whiteboard at the same time. So one AI can pick up a task, another one can finish it, another one can check
the work. And what that means for you essentially is like you could have one AI writing blog post, one
sending emails, one checking the website all at the same time, right? And if one AI gets stuck or crashes, another one can automatically take over. So
you have a full team working 24/7, but the awesome thing about it is it's AI, right? So it doesn't get tired, doesn't take weekends off, whatever.
For/goal, the AI that never forgets your mission. So you can type for/goal and then tell the AI what
you want to achieve inside Hermes agent and it just locks onto that goal and keeps working at it until it's done. Before this update, the AI would
forget what it did sometimes, you know, or it forget after a few messages. Now it just stays on track like a laser. So you give it a mission and it
just focuses on that. Your session restarts. What this means basically is like before this update if
Hermes agent crashed or restarted you would lose all the history of it. Right now it saves your work automatically and picks up where you left off. So
you don't lose hours of work and it's kind of like an autosave feature. You have her Google chat supported as well. So Hermes can now talk to you
through 20 different messaging platforms. You know for example like the usual WhatsApp, Telegram, etc.
But Google chat is in there too. So it basically means like whatever app your team uses, Hermes agent can now live inside it. Bunch of security fixes.
You've got the AI um checking its own work. This is post write lint, right? So what this means is like after Hermes agent writes any code or file, it
now automatically checks it for errors. So it's like spell check but for code, right? And why this
matters is basically you have fewer broken things that are shipped which means less time fixing the mistakes your AI made. It now speaks seven
languages. So Hermes agent now shows its menus messages in Chinese, Japanese, German, Spanish, etc. So if your team isn't English first, they can now
use Hermes agent more comfortably. You can also watch videos and understand them too. Uh this works with
Google's Geminis's AI model. So this is pretty powerful stuff. So it can watch videos etc. You can also use the text to speech and clone voice feature
with X AI custom voices which is pretty crazy. Um so your voiceovers can sound like you basically. And then you got new AI models Matt added as well.
So there's four new AI models that were added to Hermes agent which are Deepseek V4 Pro Gro 4.3
alpha which is important to know because that is a freeto use API on open router and you got Tenscent HY3 right so bear in mind Hermes is free and
open source and then you can connect it to our alpha and that is free to use too right pretty powerful stuff so basically why that matters is like
more models equals more choices and you can pick the best AI brain for each job that you want to is now
we've got a 30-day road map and step-by-step uh operating procedure for implementing this inside your business. Uh that's inside the AI profit
boardroom. But just to recap, really this is a powerful update, super useful, and uh it makes Hermes agent more autonomous than ever before. So, if
you want the full guide to this, we I'm going to post it inside the new daily updates and advanced tutorial
section of the AR profit boardroom.
So you can get the full guide right here. Uh along with new daily tutorials like you can see. Um we add video tutorials and step-by-step guides as you
can see right here. Comes with all my best training inside the air profit boardroom. It's a community just focused on number one connecting to each
other and learning and growing and making some great connections. Number two, you can ask any
questions, get help and support whenever you need it. You can post questions inside the community. We actually post all our improvements inside the
change log as well, so you can see what we're improving step by step. Inside the calendar, you can jump on weekly coaching calls, share your screen,
meet other people, ask questions about your setup, get coaching on AI agents like Hermes and OpenClaw,
etc. Inside the map, you can connect with people in your local city who are also using this sort of stuff. So, it connects you to awesome people. And
then inside the classroom, you can get access to all my new training. Feel free to get that link in the comments description or go to the
aiprofitboarding.com to get access to the AI profitable boarding.
Today we are going to be testing out Hermes agent and how to use it for free with local models with LM Studio. So LM Studio can run with Hermes agent
now which means that you can run Hermes agent and use it for free with local models. Bear in mind Hermes agent itself is a free open source model and
then if you can plug in a free LM studio local running AI model then you can use it for free as well
right so it has a a free sort of way to run it inside Hermes. So let's get started with this right now and you can see that it's quite easy to
integrate these two together. So the first thing that we need to do is run the server from LM Studio. All right. So we're going to open up LM Studio.
If you don't have LM Studio already,
you can get it at lmstudio.ai. Right? It's a free app. You can download it and then you can install local models. Right? So for example, you can set
up a local server and load a model with a local server and this means you can run local models with LM Studio. So for example, if we search inside the
model section here, we can install something like for example Gemma 4 and then run that with our AI
models, right? It's pretty simple and easy to set that up. So for example, like this. And depending on your setup, LM Studio actually tell you what's
like the best model to run. So you can see for example here it doesn't recommend using these models cuz it says likely too large. But you could run
like a more lightweight model like for example this. And the other thing is that's important to note
here is like with LM Studio you get quantized versions of the same model which means that you get more lightweight versions of the model that are
still powerful for running as a local AI. And then you can also use Hermes offline as well. That's the other benefit here. So, we can download Gemma 4
like you can see here. And this is a super small lightweight model, but I just want to show you as an
example of how you can get that set up. And then what we can do from there is we need to start LM Studio, right? Like this. So, all we do to do that
is we can go into LM Studio. Close that. Go to local server. And then we just have to run it. Right. So if we click on running now, LM Studio server
is now running locally. Right? All we did was we just toggle that bit right here. Right? So we toggle
that on. Once we've done that, we just need to load a model onto it as well. So if we load a model here, we'll just get that model from Gemma 4 which is downloading right
Then we can get this started. Now once you've started the local server, as you can see right here, we then need to run Hermes agent with LM Studio as
a model provider. How do we do that? So we're going to go to Hermes setup inside terminal here as you can see. And then what you want to do is
complete the LM studio setup from there. All right. So you see how we've got LM Studio on the model we can
select. We can go from there. Bear in mind as well you can also use this with O Lama which runs locally too. So for example we've got Gemma 4 with O
Lama uh set up locally too. You can also plug in an API key if you want to add an API key. And you can just go with the defaults if you want as well.
Then we can restart the gateway to pick up the changes. That's basically how we can set this up. Right now you can see that Gemma 4 is now downloaded.
So we can use that inside a new chat. And if we go to the local server here, we just have to load the model, which is Gemma 4. I wouldn't recommend um
using, you know, like this version of Gemma 4, but it's just as an example. And that's basically how you can get this working with uh LM Studio. And
that's basically how you can set this up. And then from here, if we go back to our terminal, click on done, we can launch Cher Hermes in chat.
Select model. Right? So we can switch model by typing forward slashmodel like so. And then we can switch to LM Studio right there. Right? So that's
how you can switch to LM Studio. And then we would select Google Gemma 4 inside the studio there.
Pretty easy to set up with Gemma 4. It also tell you like what the context window is, the provider, the model you switch to. You can have multiple
different models with LM Studio 2. Now, if we wanted to run this locally with O Lama as well, let me show you how to do that. Um, and again like
inside LM Studio, I don't recommend using that version of Gemma 4. Just wanted to show you like a quick way
um of how you can pick the model, load it in the server, and then from there connect that to LM Studio, right? That's an example. But some of the best
options that you can do here for this sort of stuff is like GLM 4.7 Flash is pretty good. And there are actually some local models designed
specifically for Hermes, right? So for example uh you can see a bunch of like news research models here. News
research is the founder of Hermes agent which means these are some of the best models you can run locally with these agents right
also you could run something like uh Quen. Quen is pretty good as a local model. I would use uh Quen 3.5 for that. And they also have some lightweight
model versions as you can see right here. So like coin 3.59B uh 35B if you've got a good setup etc. Right?
So that's basically it. And then if you wanted to configure this with Hermes um using Olama as a local model setup, you can go over here cloud.
You can plug in the API key if you want to use the cloud meth method. And then from there, you're good to go. That's how you can run it.
So just to recap on this whole process, you can now run powerful free AI agents like Hermes on your own computer with no internet, right? using LM
Studio. LM Studio is a free app you can install. It lets you download and run powerful AR models. You have to have a good setup for this to work, but
it will work on Windows, Mac, and Linux. I do this on a Mac Studio. Honestly, I prefer like cloud-based
models, but if you do want to know how to use LM Studio, this is where you can do it. And you might be wondering, okay, how does this work together?
So, LM Studio acts as an engine. It runs the AI model on your computer and Hermes engine uh agent acts as the driver. So it tells the AI what to do
and gets tasks done for you. So together they create a local fully private fully free AI agent system.
Now why should you care? Well, it's free, it's private, it's fast as well, right? And it works offline, right? So if you're like on a plane or
something like that, you can still use Hermes and you can still get the most out of it, which is pretty cool.
So, some examples of models you could use that could be like the Quen 3 series could be uh Deep Sea Coda Llama as well as another one. And if you want
to set it up, we've got a step-by-step guide right here and a full road map on how to implement it. And what I'll do is I'll put that inside the AI
profitable boardroom.
and we'll add the video tutorial as well inside there. But that's the full guide on how to use it, how it works, why you care, um how to set up, plus
a 30-day road map for implementing it into your business. Now, inside the air profit boarding, we had like new daily tutorials, full step-by-step
guides on how to use all this stuff, which is pretty amazing. You also get all of my best trainings on
like AI SEO, how to go from beginner to expert with AI automation, my best trainings here, all the systems I use inside my business, and a full course
for agencies and how to get clients as well. So feel free to check that out. Link in the comments description or go to the AR profing. This is my
community that's focused on helping you grow and scale with AI automation. Inside the community, you
can ask questions, get help and support whenever you need to. Inside the classroom, you get my trainings. In the calendar, you can jump on weekly
coaching calls and get help and support whenever you need to. And also inside the map, you can connect with people in your local city who are also
using AI agents just like you. So feel free to check that out. Thanks for watching. I'll see you on the
next one. Cheers. Bye-bye.