OpenHuman offers a pragmatic, API-driven shortcut to productivity that bypasses the current unreliability of visual agents like Hermes. It is a clear choice between immediate office utility and the raw, unpolished potential of developer-centric autonomy.
深掘り
前提条件
- データがありません。
次のステップ
- データがありません。
深掘り
OpenHuman Honest Review: Is it Better than Hermes Agent?追加:
Folks, there's this new agentic harness called Open Human built by Tiny Humans AI. They've been trending very recently.
They got about 24,000 stars on their GitHub and uh just today they surpassed Hermes agent on star history. In fact, quite a few people are saying, "Whoa, Open Human, this is the next Hermes agent, right?" Because in terms of the current history we have of Agentic Harness, Open Claw came first, then Hermes agent, and now they're saying Open Human is the next big agentic harness. Well, the short answer, guys, is no. Okay, not even close. All right, we we tried out Open Human and we understand that they're in beta, so we're trying to be polite here. A lot of stuff are broken. So, we'll go through we'll give you a review of uh open human, but we'll also tell you why they might have the potential to be a very strong agentic harness if things are in fact working. All right, let's dive right in.
All right, let's briefly go over open human in case you haven't heard of this.
So it's it's basically an agentic harness just like Hermes agent and open claw. But the biggest difference here is this is a desktop app. You can actually download this uh on Windows, Mac OS, uh and Linux. We got it already set up right here. We'll we'll look at the app later. But let's understand what is open uh human first. You know what is so special about open human? So it is a signed app for Mac OS, Windows and Linux. That way you don't need to touch a terminal to get started, which is great for beginners. Uh, and once it's running, it automatically pulls your Gmail, Slack, and Notion every 20 minutes into what it's called a memory tree. Uh, and that data lives in the SQite and Obsidian compatible markdown vault uh, right on your machine. So that way, you know, it's not in a black box that you cannot inspect. You can actually customize, edit those data. And it also has a desktop mascot with voice input, lip sync responses, and it can even join your Google Meet calls as a participant. So that's actually why they call it open human. Uh debatable. And it comes with token juice. So this compresses your contacts by roughly 70 to 80% so it stays cheap to run. And the headline feature here which is the potential that they can be a very strong agentic harness is their one-click oath for over 100 apps including GitHub, Jira, Linear, Notion, Slack, uh Google Docs, you name it. So, this is the piece that we're going to be focusing on here in this video. All right. Now, we're going to look at the app. And this is the biggest frustration that I had today. It was so buggy. They are in beta. Remember this, okay? If you're a nice person, you can forgive them for this, but uh you'll have to sign in to your Google. And once you signed in, you want to click open uh open human right here. And it seems like, right, like you have to sign in again. You're stuck in this loop. But when you're at this phase, once you've signed in, you click that button, you click uh run locally.
Recommended. It's safer to run on the cloud on a VPS, but you know, for the sake of this video, we're just going to run locally first. Uh, okay. Now, yeah.
So, if that happens, just sign in again.
You know what? I'm just going to record this from my Mac. All right. So, I'm on my Mac now. I'm doing it again. I signed in. Uh, then the trick is just to click run locally. And there you go. You want to see this one uh get started. And uh yeah, this is also another barrier for Open Human is you need credits in order to use OpenHum. Uh I'll show you later, but they have a pay as you go structure or a subscription tier. So yeah, that's that's a big barrier there. But they do have $1 free credit to try out. So you just want to click that one. Continue with simple. And here again, another UI bug. It will take a while for some of you. Oh, okay. This one was fast. Okay, so once you're here in the app, right, this is the open human mascot. It's a very cute little guy, but you're not going to trick me, little guy. I mean, it just looks like Open Claw, but with a I don't know what that is. A haircut. Uh little little blob. Halo. Uh but most of your work will be done here in the chat.
So, it's it's a very simple UI. You you type O. Okay, I'm going back. Why did it bounce me back to the home? So, in practice, this is where you work with your agent. Uh, but as you can see, I'm trying to chat with my agent, but there's this error, uh, system message error there. Uh, and it's annoying how it's not showing me there. It's just one enter, and it only briefly shows. It's saying realtime socket is not connected.
Responses cannot be delivered without a client ID. All right, that's because the open human core did not do the handshake with the app. So, uh, there's no client ID and there's nothing I can do about that. I hope the team will address this.
But let's get to the focus of this video, the point of this video, which is the connections, the integrations. This is I'd say the biggest strength that open human can potentially have if it works uh over Hermes agent because you can have native API connections of these biggest apps to your agent where they can actually use these apps on your behalf. And of course you can do this with the computer use of Hermes agent which we did a full guide about where you can get your Hermes agent to access your app and use that app like we did for trading view. Uh but the biggest draw down of using computer use through your agent to harness is very timeconuming because you have to give approvals for every action that they take. And another flaw of computer use for Hermes agent is most of the time they would just see pixels and they would guess where the UI elements are even though you enabled screen recording in your system settings which theoretically means they can actually see every pixel going on uh on your on your device. Most of the time they are sort of guessing where the UI elements are. So it must be manually scripted or learned. So it's a very long process to perfect your Hermes computer use and often your Hermes agent on computer use would break if a let's say a website updates is it's uh CSS right and it has to scroll and click through pages one by one like a human would uh but with the native API integration this is what you need because it's native ooth integration uh and they're using uh composio which is structurally better than the screen and mouse uh computer use approach. Uh and with Composeio, it's bundled 118 OOTH integrations. So either you know Open Human fixes their app, gets out of beta, ships it officially, so it's actually not broken or we wait for Hermes agent computer use to be much better. That way they guess less. All right. uh and they are much more adaptive and they actually see what we see and interact with the apps that we have. So that's basically what I think that open human can potentially have advantage over Hermes agent. Also a couple of comparisons to note here is open human is a desktop app first. Uh even though they have the option for you to run locally or on your cloud Hermes is first and foremost a CLI and server tool. So you can run it on VPS or in docker or in local. So open human wins on setup speed. It's just you know uh a desktop app download which by the way I think maybe you know at the maybe the endgame is all agentic harnesses must be an app right that sort of makes sense because if you want mass adoption of agentic harnesses then you have to have an app cuz people most people you'd be surprised. I mean you guys are watching this now you're hella smart. no glaze.
But you'd be surprised how many everyday people, most of my other friends are like, "Ew, terminal, what is that? Are you like a hacker man?" Right? But yeah, if they see an app, then they'd be more inclined to try that agentic harness. Uh so yeah in terms of setup speed open human wins but in terms of depth you know the accidental stuff that you need to get stuff done Hermes agent right has natural language cron scheduling isolated sub agents you got the canban orchestrators you got the skills bundles uh even though the computer use it can be a bit flawed especially in long workflows for much complex workflows I would say there's still very usable for easy tasks. Okay, so the trading view example I showed you guys on that video was a bit of cheating because you can see I caught this postedit by the way. I was like, "Oh no, Hermes didn't actually check my Trading View desktop app. I saw in the terminal they were actually fetching Coin Gecko API data." So bit of a cheat there, but it's still uh feasible. You can still use computer use for your Hermes, but a lot of work to do there. So the TLDDR here is if open human becomes better in let's say a couple of weeks or a couple of months and it actually becomes really solid. It just comes down to one question. If you are considering should I use open human or is it Hermes agent it just comes down to if you want ooth connectors for apps or if you want to bring your own key uh integrations uh a lot more customization right you can even have both that's perfectly fine but at the end of the day what who is open human for right now is more so for office workers right so um you see all of those apps the the native API integrations are more so for office work. Uh even in their video they actually have a YouTube channel the they did a demo guide where the guy says this is perfect for white collar workers.
White collar workers. So office workers.
So basically if you just want your agent to have immediate access to those uh office apps uh emails, docs, chats, whatever then yeah open human. But if you want full customization, you want full autonomy of your own project, then of course Hermes agent. And it's too bad they don't have an official trading view integration as well. When I first heard about open human and and know that they have this native API integrations for hundreds of apps, at first I was thinking, oh okay, yeah, that would be the big advantage there if I can just directly connect to Trading View and get my agent to use that Trading View. But no, so far no. So, it has to be Hermes Asian for me in my personal case. But now you know exactly if you need open human or not. So, if you find this video helpful, smash up that like button. Subscribe to our channel to follow for more updates and more guides. My name is Ron. Shunning out.
関連おすすめ
OpenHuman VS Hermes AI: Who Wins?
JulianGoldieSEO
285 views•2026-05-29
BREAKING: Microsoft’s New Image Generating Model Beat Out GPT 1.5 and Nano Banana 2
aimmediahouse
122 views•2026-06-03
Long-Running Agents — Build an Agent That Never Forgets with Google ADK
suryakunju
142 views•2026-05-30
This computer is made from real human brain cells. And you can buy it.
Talktmsmedia
3K views•2026-05-28
I Made the Same Anime Fight Scene in Every AI Video Generator
NobleGooseAnime
295 views•2026-05-30
Nvidia Bets Big On AI PCs | New Chip To Power Windows Laptops | Technology | AI Updates | N18S
cnnnews18
3K views•2026-06-01
I Tested NEW Opus 4.8 on Four Projects (Updated LLM Leaderboard)
AICodingDaily
298 views•2026-05-29
3D Platformer Update - NO CAPES
SolarLune
294 views•2026-05-30











