**Julian Goldie** (0:00)
Today, we have a brand new update from Hermes as they've released their new desktop. And you can see it installed right here. So this is a free desktop app that comes directly with Hermes Agent. You can see it's available on Windows, Linux and Mac OS right now.
Lives everywhere, it's got persistent memory, focused automation as well. You can actually browse the web with it. It's got isolated sandboxing and you can multiply your tasks with Hermes Agent. So today, I'm gonna walk you through exactly how to use it, what it is, why it works, what the point of it is and some cool stuff plus what it looks like. So let's have a look. This is Hermes Agent with desktop. Now, if you want to get it installed, you can just go to their website here and then you can download it. Now, you actually automatically detect where to install it to. So for me, for example, it recommends using this on Mac OS. Now, you can see an example of how it's been used here. So you got all of your conversations in one place across multiple platforms. So that's pretty useful. And then you can ask it to go off and start doing things as well. It can also talk cool tools. So for example, what you can see is that it can use the agentic features of Hermes Agent inside the desktop app. And then we can go inside here and use it directly. It was pretty quick to install, to be fair. And you can also pin sessions. So we can see all of our previous sessions here and then we can pin stuff too. Then we can go to a new session. We've also got some options for our settings. So inside the settings section, you can actually change the context window. You can change which model you're using with your agent. So you can see I've plugged in Minimax M3. You can change the models that you use for each. So for example, like session searches or MCP and that sort of thing. And then you've got different options, for example, like the chat, you can actually change the personality. So you can have helpful, concise reminds. It feels like they're trying to do something similar to what called desktop have done, but the Hermes agent. And then you can switch between these different options right here. You've got the appearance options. You can switch to dark mode, which I think is probably better to be fair. And then you can change the theme of this as well. So we can change to midnight, for example. Over here, we've got the workspace options. And these are just some ways that you can basically customize this as much as you want. We have the options, for example, for the user profile, persistent memory. Then we have the context engine. So how that works, basically, whether you want it to automatically compress the conversations that you have so that it doesn't run out of context window, depending on what model you have. If you're running a smaller model with a 256k token context window, then that's probably better for you. And then you've got the threshold here in terms of like when should it compress. Over here, you can change the voice. So we can select which voice we want to use. Now, you can switch between what type of provider you want as well. So you could use like AXAI, you could use ElevenLabs, you could use OpenAI for your voice responses. You could actually switch off speech to text if you want to. And then additionally, you can change the models down here too.
So one thing that would be pretty cool with this is like you can actually change the ElevenLabs voice to like my own voice that I've plugged into ElevenLabs. And then it could respond with my own voice, which is pretty weird and interesting at the same time. So over here, you can change the gateways, you've got APIs as well. And so you can select your APIs, switch between APIs, change the models. And then there's APIs not just for the models inside Hermes, but you can manage the APIs inside the tools you use as well. So one thing that's really good about this is like, if you've ever done the old way, which is you go into terminal here, you get Hermes Agent running like, and you can't really customize anything, right? And also 99% of people are just not going to use the terminal. It's not very useful. It's kind of too technical for most people. And also you can't manage the models, the APIs. You can with some terminal commands, but most people don't even know the terminal commands in the first place, how to type them in. So it is better to have a desktop like this. It's great for, you can see what they're doing, because they've already become the most popular AI agent in the world if you look at OpenRouter and the usage of which API is being used across these agents. And so if they've made it even easier and less technical to use AI agents like Hermes, then it just grows, it expands, more people use it, right? So that's obviously what they're doing here and it's smart, it's smart. And it also makes it just easier and more fun to use your AI agents. Now you'll see over here, this is a section where you got basically your artifacts, so things you have created. Now one thing that I will say about this is not great is like you've got your files and everything else like this, but if you actually click on them and you want to view something like this and preview it, you'll see that it opens up, but it will open up in like this sort of visual studio code here, or whatever you use as default to open up like a.md file. The problem with that is that you can't really preview the things you've done properly. So I'll give you an example. If we go over to my agent OS system, if I'm like trying to preview things that I've created, I can easily just go in the studio, I can see the videos, I can see the boys, I can see the images I've created here, I can switch between Grok and Minimax, which is pretty cool.
11 more minutes of transcript below
Try it now — copy, paste, done:
curl -H "x-api-key: pt_demo" \
https://spoken.md/transcripts/1000651996090
Works with Claude, ChatGPT, Cursor, and any agent that makes HTTP calls.
From $0.10 per transcript. No subscription. Credits never expire.
Using your own key:
curl -H "x-api-key: YOUR_KEY" \
https://spoken.md/transcripts/1000770974503