NEW Hermes AI Voice Agent is INSANE! artwork

NEW Hermes AI Voice Agent is INSANE!

AI News Today | Julian Goldie Podcast

July 14, 2026

NEW Hermes AI Voice Agent is INSANE!
Speakers: Julian Goldie
**Julian Goldie** (0:00)
Hey, how's your day going? Quite smooth, sir.

**SPEAKER_2** (0:04)
Efficient and pleasantly uneventful, just as I like it. How may I assist you today?

**Julian Goldie** (0:09)
Yeah, I'm doing great. Could you give me some ideas for like keywords for SEO?

**SPEAKER_2** (0:14)
Certainly, sir.
A reliable approach is to mix a few types. Broad phrases that describe your main offering, long-tail phrases that match specific customer needs.

**Julian Goldie** (0:24)
Actually, stop, can you just open up Google? Perfect.
Can you tell me a joke now?

**SPEAKER_2** (0:31)
Certainly, sir. Why did the scarecrow win an award?
Because he was outstanding in his field.

**Julian Goldie** (0:39)
Can you teach me a couple of phrases in Italian?

**SPEAKER_2** (0:43)
Certainly, sir. Here are two handy ones. First, buongiorno, meaning good day, pronounced buongiorno.
Second, per favore, meaning please, pronounced per favore. They'll serve you splendidly in most polite situations.

**Julian Goldie** (1:01)
Fantastic. I'll tell you what, could you just build me out a website for my SEO agency?

**SPEAKER_2** (1:06)
Very good. Sir, I'll put together a clean professional agency site concept now.

**Julian Goldie** (1:11)
So I'm just going to move that for now, but you can see here that it's building out the setup. As you can see, I'm going to switch that off now. But essentially what this can do, this is Hermes Apollo, which is a custom made voice agent that would build into our agent operating system. We can put it in wall mode, so we can have it in full screen. We can see everything that we've spoken to it about. We can see the latency details over here, and it looks really beautiful. We can also switch real time on and off, and we can have a wake word, which means, for example, we can say, for example, Apollo and then it will wake up automatically instead of having it listen and reply to everything that I want whenever we want. Over here, you can see that we've got the full script of us going back and forth, and then we have everything that we've built with the AI agent as well. So it is very, very quick to reply, the fastest I've seen, and we're using something called ChatGPT real-time API to build this system. And this system is something I call the Apollo Voice Command Agent. So what we have here is Hermes plugged into my Apollo Voice Agent, which is basically a butler that lives in your agent OS, and it can run your computer by voice, right? So you can talk to it, it can act, it can open up websites, it can open up apps, it can build real working apps, games, tools, apps, or whatever you want, whilst you watch hands-free in real time with no clicks, and there's no lag on it as well. So here's exactly how it works.
Now, the first thing that I would say here is that there's a bit of a keyboard problem going on right now with AI. So we all know what we want with AI, but typing and everything else is a bit of a drag. And then also, if you like doing something else in the background, you know, quite often we're coding with like multiple terminals, multiple agents and that sort of thing. With this system, really, the way that I would look at it is that you can run your whole day without having a keyboard. So think about like the last hour at your desk, for example, maybe you wanted something, maybe you wanted to create a tool, check something, spin up a quick website, whatever it is, and you have to type it.
Whereas with this system, you don't need to worry about that, right? So when you actually do want to make something, you're not staring at a blank editor with your hands on the keyboard. You are just instantly commanding your agent to go off and do stuff for you. Now, some people say like, you know, voice agents, aren't they just a gimmick? But this one is actually built inside a system that you can use every single day, right? So we've got the HNOS, we've got all of our different tools over here. So for example, we have GPT 5.6 plugged in to call to code, we have Hermes agent with the chat over here, we have Hermes Oracle that's refreshing every 24 hours and coming up with the latest news headlines and researching my competitors and the latest keywords inside my niche. And then we have Hermes Apollo that can reply to me in real time.
And so this is a voice, a fax brain and a pair of hands. And there's three pieces to this. So number one is your voice, the input. You just talk out loud like you talk to an assistant in the room. There's no wake word to remember, no button to hold.

9 more minutes of transcript below

Feed this to your agent

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/1000776692132