Hermes Agent V0.12 Just Changed AI Agents Forever!

Hermes Agent V 0.12 just dropped. It's a free new update and it just got about 100 times more powerful. Your AR agent can now clean itself, think for itself, and run meetings for you all without you touching anything. It starts faster, it gets smarter every week on its own, it joins your Google Meets, transcribes everything, and follows up after.

Even if you miss a call, your agent can go instead of you. It can even control your Spotify now, play music, search songs, manage your playlist, all through your AI agent. And all of this is free. There were 213 people that worked on this update, 360 bugs, 1096 improvements made.

This is not a small patch, this is a completely different AI agent. And in the next few minutes, I want to show you exactly what's new, how to update in literally one command, and what you can do with it today that you couldn't do yesterday. Let's get straight into it. Today we have a brand new release from Hermes Agent V 0.12, the curator release.

I'll show you exactly why that is and how to update now. So if you want to update your new Hermes agent to the latest version, you're going to go inside your terminal and you're going to type in Hermes update, like so. That will fetch the new update and begin to install it right now, as you can see. And I'm going to run you through exactly what this release means and why it changes everything for you today.

This was literally just dropped. And one of the biggest updates here is the autonomous curator. So you can actually use this command, Hermes curator, which runs as a background agent on your gateway. And it runs every seven days.

It grades your skill library, consolidates related skills, prunes all the bad ones, and writes reports as well on exactly what it did. So it can clean up and manage all of the skills that you've set up so far. And we've also got a new self-improvement loop that has significantly upgraded as well. Once you've upgraded and updated, like so, to the latest version, you're then going to run Hermes like this.

And you can see we are on the latest version here. Now if you want to run the curator, which is the new part of the update, the biggest update right here, and it also ranks your skills by usage, you can go inside a new terminal tab. You would type Hermes curator like so, and you can see it right here. So if we do Hermes curator run, we can trigger a run, or we can actually check the status, right?

So if we check status, we can see what it's doing. If we do Hermes curator run, that's now running a review of all of our skills and using the curator in the background. Now that is running in the background, which means it's running autonomously. And then if we want to check in and see what happened, we can just use the terminal command, check Hermes curator status, and we can see a full setup on that later.

There's also a bunch of new updates. So for example, 50% faster, 57% faster on the startup as well. And we've also got skill integrations. So it's using Comfy UI.

I've actually got a tutorial on that, that I posted yesterday on how to set that up. But you can have that running locally, and it can use local models or cloud-based models to generate stuff as well. We have LM studio, which is now upgraded to work with Hermes. So what that basically means is you can run local models.

So if we open up LM studio here, this is a powerful way to run local models with Hermes agent or just in general. And you can actually search through and see what's just come out. So for example, we've got Nemetron 3 inside there. We have minimax m 2.7, Quen 3, Gemma 4, etc.

You can see all of the new models that have just been released recently. Even DeepSeek v4 is inside LM studio. And that's a free way to run local models with your Hermes agent. Yumbao has been added as a messaging platform.

It's also got tools for Spotify. So this actually links to Spotify now too, which is pretty crazy. If you want to set that up, you can use Hermes off Spotify, link to Spotify here, and then you can sign in and use that for your Hermes agent, which is amazing. I never knew that existed.

We also have a Google Meet plugin, so you can join, call, transcribe, speak, and follow up with Hermes agent, which is pretty powerful. So your agent can now join Google Meet with you. Now I have tested this with OpenClaw and it was very, very buggy, but I know that Hermes agent is way more powerful and way smoother. Honestly, if I'm going to choose between the two, I am using Hermes right now.

So just bear that in mind. And so what this can do is actually it can join meetings, it can enable captions, it can set up a transcript file, and all you do is you use this command. So you put meet underscore join with the URL of the agent that you want it to join. And you can run Hermes meet install to set this up, right?

So if you want to use this plugin with Hermes agent on Google Meet, you just use Hermes meet install and then it can run from. After that, you use the Hermes meet join and it can join your meetings, transcribe them, fill you in on the details. I mean, you could even miss a meeting and then your Hermes agent can join for you and give you the details later on. So let's talk this through and I'll guide you through everything that's changed.

So the big idea here is, you know, what changed? Well, the number one thing you need to know is the curator, right? Before v0.12, Hermes agent would learn new things and save skills, but nobody cleaned up the mess, right? So old skills would pile up, slow things down, and then that would waste a lot of tokens as well.

Now Hermes has a built-in janitor called the curator that runs every seven days automatically. The curator looks at all your save skills and asks is this actually useful? And then it will delete the old ones, merge the similar ones, and keep the useful ones, which means your AI system gets smarter over time without you doing anything, because this is autonomously running in the background like we showed before. And if we want to just check where that's up to so far, we can run Hermes curator status like so, and we can see where it's up to.

And we can see the last run, the last summary, what it did, how often it runs, anything it made and edited. Now it also starts 57% faster. Nobody likes waiting around. I have noticed, for example, OpenCore has got a lot slower when it opens now, and I think that's because it's using plugins it shouldn't, but this basically means that the startup time has been cut by more than half with Hermes, so you can open up Hermes and it's ready almost instantly, right?

And this is done by making it load only what it needs when it needs it, so it just runs a lot more lightweight than it did before. Spotify is built in, so you can tell Hermes to play music, search songs, control your playlist. It logs in via the secure login, and then there's a setup wizard that walks you through connecting it in minutes. We have the Google Meet integration, so Hermes can now join your Google Meet calls, it can transcribe what people say, actually speak on the call, and follow up afterwards, which is huge, you know?

And then you've got two new chat platforms, so it now works on Microsoft Teams, which is the 19th messaging platform it supports, and also Tencent Yunbao, which is a major Chinese chat platform, that's number 18. And you can control your AI agent from pretty much every messaging app you have now. Now we also have a better command line experience, so it renders latex, aka math equations, properly. You can also delete messages you're typing with Ctrl and X, so if I'm typing like so, and then I press Ctrl and X, it can actually delete messages.

Some security add-ons too, four new providers too, so you can use Hermes with four more AI services, which is GMI Cloud, Azure, Minimax, and Tencent. LM Studio is also supported. Comfy UI and Touch Designer are now in there as well, so Comfy UI lets you generate AI images and videos. Touch Designer is a tool for visual art and creative coding.

Both can be used as optional add-ons, and they come bundled in by default when you set up the new version of Hermes. We also have a remote model catalog, right? So new AI models now show up in Hermes without needing to update. Open Router and News Portal model is supported.

So what this basically means is like, if you have a new model that gets released on Open Router, you don't need to update Hermes to get that new release, right? The model catalog automatically updates without even using the news update. So there were 213 community contributors who helped build this open source release, 19 messaging platforms, 57% reduction in startup time, 360 bug fixes, and 1096 commits made since the last version, and 550 pull request merge, basically improvements. And if you want to get this set up, we've got a full setup guide here, along with a 30-day roadmap on exactly how to use this stuff.

This is all available inside the AI Profitable Room. But just to recap everything you need to know, Hermes Agent is a free AI tool that lives on your computer and actually does things. Version 0.12, released 30th of April, is called the Curator Release. The biggest new feature is an autonomous background agent that cleans up your AI skills every seven days automatically.

It now starts 57% faster, works on 19 messaging platforms, and supports Spotify and Google Meet natively. It's got four new AI providers added, including Azure, AI Foundry, and Minimax. 213 community contributors helped build this release, with 1096 code changes. To update, you just type in Hermes Update in your terminal.

To run it, type Hermes, or Hermes, and insert your task here. And the self-improvement loop now safely grades and updates itself without you touching, or it touching, your shell or web. So this is basically a much more powerful autonomous AI agent that's self-improving. So thanks so much for watching.

If you want to get the full guide and 30-day roadmap on how to use this update, I'll put it inside the AI Profitable Boarding. Link in the comments description, or go to the AIProfitableBoarding.com. Inside this community, we have new, daily, advanced AI tutorials. We do full step-by-step guides.

You can get all of my best training. This is a community focused on helping you save time and grow with AI automation. And inside the community, you can post questions, get help and support whenever you need it. You can also jump on the weekly coaching calls, ask questions, share your screen, etc.

Inside the map, you can connect with people in your local area and city who are also using AI agents. And then inside the community, you can get help and support. And I answer daily Q&As inside there too. So feel free to get out.

Link in the comments description, or go to the AIProfitableBoarding.com. Thanks for watching.