# NEW FREE OpenClaw Update + Alibaba's Copaw vs OpenClaw + ClawX Desktop App

## Метаданные

- **Канал:** Julian Goldie SEO
- **YouTube:** https://www.youtube.com/watch?v=SVwLd9FXL6E
- **Дата:** 03.03.2026
- **Длительность:** 53:24
- **Просмотры:** 4,612
- **Источник:** https://ekstraktznaniy.ru/video/10762

## Описание

Want to make money and save time with AI? Join here: https://www.skool.com/ai-profit-lab-7462/about

Get a FREE AI Course + Community + 1,000 AI Agents + video notes + links to the tools 👉 https://www.skool.com/ai-seo-with-julian-goldie-1553/about

In this livestream, I cover the new FREE OpenClaw update, pit Alibaba's Copaw against OpenClaw to see how they compare, and introduce ClawX - a brand new free desktop app for OpenClaw AI agents. If you want to stay on top of the AI agent game, this is the stream for you.

## Транскрипт

### Segment 1 (00:00 - 05:00) []

OpenClaw just dropped its new free update for March the 2nd and it's packed. We're talking AI that can now read your PDFs instantly. AI agents who can talk to each other and hand off tasks without you lifting a finger. Realtime responses stream live so it actually feels like conversation inside Telegram. Plus over 100 security fixes so your setup is better than before. This is not an upgrade. This is a completely different tool to how it feels like last week. And here's the craziest part. is completely free. You can get updated and set this up today. We're going to look at every single one of these new features together right now live so you can see exactly what this thing can do and whether it actually lives up to the hype. If you've been waiting for one AI tool to run your entire life, this might be it. Let's find out. So, we're going to get straight into this. You can see the update right here. And this is the update. Now, the first thing to note here is the Telegram um streaming example of the messages, right? So, you can see the messages and how it comes in real time. uh like this inside the video. So, if you want to see the full change log, here are all the new updates like you can see. So, a bunch of changes, a heavy list of changes as you can see. There must be what, at least 50 different changes added to this. Um, and there's a bunch of new updates here. Now, if you want to update this automatically, what you can do is just go to your config file here. Scroll down inside open claw if you've got this running. And then you're going to make sure that you have auto update enabled over here. Auto update. Right now, once you've done that, we're going to go over to the chat and we're going to click update now. You can see the notification here. So, it says the um 2nd of March update is currently running the 1st of March update. So, we need to update. We're going to click on that update now button. Sometimes this breaks, sometimes it doesn't work. We're going to figure out what happens together and test it out and we'll see how it performs. So, it's beginning to update now. Sometimes the gateway breaks. I will warn you now. But if it does, it's okay cuz I'll show you exactly how to fix it along the way and how to get this working. Now, let's talk about the update whilst it's beginning to auto update itself. Um, and we'll talk you through some of the changes and how it happens. So, if you've never heard of OpenClaw before, I'll keep this short, but you're about to be very happy on what you stumbled on. This free tool got one of its biggest updates ever. Um, and it's basically like a really powerful sort of AI that runs locally on your laptop. Um, and it lets you, for example, connect AI systems, automate tasks, build custom workflows, and do all of this. And it's completely free and open source to use. You can even connect it to for example uh Olama or Quen 3. 5 and run this locally without any API costs as well which is pretty awesome. So once we've done this we can see it. Now I'm just going to check the terminal and make sure that it's working. Check the logs inside the terminal. When I actually click update now it doesn't seem to be but let's see if it's working. I'm just going to say are you working? Sometimes it just takes a little while to respond. So, I'm going to see what happens uh when we click are you working. It might have just broken again. So, let's test it out and then let's talk about what changed in the meantime. So, now you get this sort of live uh messages inside Telegram, right? So, what that means is that you get the responses instantly and you can see them as they go along inside the chat, which is pretty cool. And then also you can basically when your AI responds inside Telegram, the text now appears word by word in real time just like watching someone type. So here's the example, right? So this is an example of an open call responding and someone said like send me a 30 line haiku and then it started to produce the haiku in real time right here. Right? It's pretty cool because then you don't have to wait for the whole message to load. You just get it um good to go. Right? And so if we go back inside the chat now I'm just going to say are you updated to the latest version because you can see already that it's getting buggy inside the chat here. So, if we click on update now, it just it clicks updating and then stops working, right? [snorts] Um, so I'm going to see what it responds with now and just make sure it is working on the right version. If not, we might have to run a terminal command to get it running. Um, or we could just uninstall it and then reinstall it again using the newest updates as well. Let's see what we got inside the chats here. So, Sakura says, "What is the best model to run locally and get results? " Um, so I would recommend checking out Olama Quen 3. 5. GLM 4. 7 Flash is a cloud model, but it runs locally with OAM, if that makes sense. So it runs via OAMA, but it's a cloud API. But if you want like a fully um local way of running OpenC, then you can use OAMA 3. 5. I actually did a video on it earlier this morning and created a full guide. You can check that out inside the AI profit boardroom. So if you go to the open claw plus aquen 3. 5 plus alarm just dropped some smaller models that are faster to run locally

### Segment 2 (05:00 - 10:00) [5:00]

but still very powerful um and those were just dropped a few hours ago. So I've got a full guide and tutorial on how to set that up as you can see right here inside the air profit boarding. So let's go back to the update details now. And if we go back to open claw you can see that it says we're actually ahead. So, it is running the March the 2nd release. There does seem to be some sort of bug inside the update section. So, like when I click this, it doesn't respond, but you can see here that it's running the March the 2nd update, which is great. Um, so it's good to go, right? I'm just going to make sure. Are you I'm going to say, are you running Let's just copy this. Are you running the update? If not, can you run can you update to the version? All right, the March the 2nd version. So, let's just see what it does in a second. If not, what I'm going to do is just I'm going to grab the URL from the change log at the top here, and then we'll plug that back into OpenC claw. Um, so now it's beginning to check with OpenClaw to see whether it's done or not. In the meantime, let me talk you through what the other updates are. Right? So, one of the other cool things here is ACP agents sub agents are now on by default. Now, if you're wondering what ACP is, ACP stands for agent communication protocol, right? It's like a walkie-talkie system that lets multiple AI agents talk to each other and hand off tasks. Now, bear in mind with ACP, you can even, for example, um you can even use something like uh clawed code with ACP if you've never seen that before. And this is switched on by default. Now, I'm sure you would have to allow permissions to get it using claude code, but you can see an example of how this works. And this means it can basically run like claude code locally for you and a bunch of other sub aents, right? So, your agents can build teams, work together, um spawn sub agents, that sort of thing, right? So, if you're wondering, okay, what changed? Well, previously you had to manually turn this feature on. Now, it's enabled automatically when you set up open claw. Uh, so you could have like one AI agent that handles custom questions, another one that handles scheduling. With ACP, they now automatically pass tasks to each other without you having to be middleman, right? So ACP dispatch now defaults to enabled. Um, and if you're upgrading from the older version and you specifically don't want agents talking to each other, you can manually turn this off with this command right here as you can see. Right. So let's go back to open claw. See what we're doing. And it's like, yeah, your open claw is already on the 2nd of March. So, this is interesting because the button at the top doesn't work, right? That it did update, but it doesn't tell me. Um, so just make sure it is running inside the chat if you want to make sure, right? And if it's struggling, all you do, I've done this before, if it breaks, is like you copy the URL over here, you go back to openclaw, you paste it in the chat, and you say update to this version, right? Um, so that's how to update, how to fix things. If you see any bugs or errors like I've just seen, then you can fix it quickly using this method. If we go back to OpenClaw and the updates now, the other thing is native PDF tools. So your AI can now read your documents. Um, so what changed is basically OpenClaw now has a built-in PDF tool that lets your AI assistant read, analyze, and work with PDFs, right? And then the supported AI providers include Anthropic and Google with automatic fallback for other AI models. What this means is basically like you can upload a PDF, the AI can read and summarize it. You know, you could drop in like a research paper, the AI answers questions about it. You can send a client proposal, the AI helps you actually edit it as well. And there are configurable limits which mean you can control how large a PDF can be when you upload it how many pages get processed and which AI model handles the reading right um the next one openclaw config validate right so you can stop breaking things before they break if that makes sense so this is quite an interesting one what changed is there's a brand new command it's called openclaw config validate right and what it does is before you start up your entire openclaw setup you can now check if your set settings are correct first, right? So, basically, it's like a spell checker, but for your AI configuration on open core, it can catch errors before they cause problems. Now, this is useful for you because, you know, like one typo in a config file uh used to mean like hours of troubleshooting, and I think we've all been there with openclaw. Now, you can run one command and it tells you exactly what's wrong and where, right? And you can even run openclaw validate JSON to get a clean structured error report. There's some other cool stuff as well. So, for example, one of the most popular messaging apps in Vietnam, Zalo, used by over 70 million people, is now added inside OpenClaw 2. And there's 100 security fixes. You can see a list of all of them right here. Right. Um, some

### Segment 3 (10:00 - 15:00) [10:00]

other cool stuff that didn't make the big headline. So, this is the big headline inside the tweet in terms of what came in. Um, but if we go back to the updates, some other stuff worth knowing. So, you can use locally hosted AI for memory search. Now with Oama, you can attach files when spawning sub agents. You can reload the config. You get transcripts of what the AI agent actually heard if you want that. You can do Docker health checks, which means you can just check um if your Docker if you're running it via Docker, you can check if it's healthy or not, right? Um and then you got funny startup messages as well. So, it's called uh plug-in banner taglines. And basically, this means when it starts up, um it gives you funny messages. That's essentially it. So, this is basically how it works. Um and you know, you might be wondering, okay, who is this update for? So, let me talk you through that, right? So, if you're a business owner, then the PDF tool and the Telegram updates are going to give you like better customerf facing AI bots, right? If you're an e-commerce seller, then multi- aent ACP coordination automates more of your workflows. If you're a content creator, the Telegram uh updates with live messages just makes AI interactions feel more natural for you and your audience, especially if you're showing this stuff. And then it can also use alarm memory and it's got more security patches as well. Um, if you're based in Vietnam, obviously it's using Zo now. And if you're just an AI hobbyist, then there's 100 extra fixes, which means things actually work better now. And that's basically it. So that's how to update. That's how to use it. You can see it's working quite nicely here. Pretty easy to set up. Even for example, we uh updated the CLI as well directly using this too. And you can now use ACP agents. So just to recap, the too long didn't read. Open claw is a free tool that connects AI models to messaging apps and automates workflows. Version the 2nd of March is one of the biggest updates yet. The six big headline features are Telegram live messages, ACP sub agents on by default, native PDF tool, the config validator, uh ZLO plugin, and 100 security fixes. If you're struggling, um you know how to fix it as you've seen today. if you've seen any issues and if you're new, this is one of the best free AI automation tools available right now and this update just makes it better, right? And the other thing I would say here is like they seem to be prioritizing security in each of these updates. So that's basically how to get OpenClaw, how to use it, how to set it up. If you want all the links, we've got it inside the video notes inside the air profit board today. This also comes with a six-h hour master class on OpenClaw and another three-hour one as you can see right here with full courses and setup guides and everything else and all the best use cases. We're constantly updating this community with all the new updates. You can see we've actually added one, two, three uh six updates today on um the newest updates. Yesterday we covered mission control for openclaw, how to use openclaw with uh clawed code as well. We've also got other stuff on like open call versus alternatives like accomplish. I've pretty much tested like all the big alternatives to open call. Um that's basically how to use it. Now if you want to check this out, this is the AI profit board link in the comments description or just go to the aiprofitboard. com. And if you want to learn how to save time, scale, and grow with AI automation. This one of the best places inside the community. You'll get access to 2,500 entrepreneurs who are doing similar things as you. You get a daily accountability group so you can check in, post your goals, and stay 100% focused. You can ask questions to get help and support inside the community, too. And I also post weekly updates, right? So, I actually tell you what's worth paying attention to and what to ignore. As you can see right here, we added a new weekly update today. Um, I usually spend about 60 hours testing this stuff each week. And so, like, I just condense that down into like here's the stuff to figure out, here's the stuff you should try, like top three things, and then here's all the other stuff you can ignore. And if you're wondering, okay, like are people getting results with stuff like this? You can actually see we have over 133 pages of wins from happy people inside our school communities, right? So, you can see all the wins, all the people getting results with this stuff. It's pretty much an infinite scroll as you can see right here. Um, just loads of people getting awesome results, which is great to see. Uh, inside the community as well, you get four weekly coaching calls. You can jump on these video coaching calls, get help, get support whenever you need to. You can meet people in your local area. Just zoom into the area that you want. And then also inside the classroom, you can see that we have a fiveweek master class that takes you from beginner to expert with AI automation and teaches you how to build your first AI agent in under five minutes. Additionally, you'll learn all of my best playbooks I personally run for my business. So, for example, like AI avatar videos, shorts, Instagram, newsletters, X, and everything else. We also have a full course for agencies. So, if you want to learn how to get more clients, you can check that out here. [snorts] If you want to get the latest updates, just check out this section. You can see we

### Segment 4 (15:00 - 20:00) [15:00]

update it every day with the most important updates. And then, if you want to watch about the coaching calls, you can do that over here. If you want to learn how to rank one with AI SEO, you can check that out. And also, if you want to learn how to grow a YouTube channel based on what's working for me with AI, there's a training on that, too. So, let's just check out the questions. I think we got one more from Naim who says how to get Kimmy K 2. 5 for free or the miniax 2. 5 right so there's a few different ways let me I want to talk to you but so what I would say one of the easiest ways you can do this right is you can actually go to um in inside the onboarding for openclaw you can select miniax right and if you do the oorthth then you automatically get opted in for free uh seven days for free as a um free trial for miniax coding. Right? So if you select miniax on the onboarding for openclaw, then you get 7 days to test out miniax for free with open claw directly. Right? Automatically um subscribes you to the coding plan. So that's one option. The other option is if you want to get Kim K2. 5 for free, you can do that with OAMA, right? So you can check out Kim K 2. 5 on O Lama. We actually have full guides on both of these things as well as GLM 4. 7 flash. So if you type in inside the air profit boardroom, the search bar GLM 4. 7, then you can see how to set it up inside this tutorial, which is a full video guide plus a step-by-step process that talks you through it. Right? If you want to learn how to set up Kimik 2. 5 with free, then you can check out this video guide and the full setup instructions right there. So whichever setup you prefer is waiting inside there for you. And then also if you want to check out how to set up Miniax M2. 5 then you can check this out directly inside here as well. All right. So we talk you through exactly how to set that up with a video tutorial and a full guide as you can see. Another question is did your open claw break when you updated? So today it didn't break although it does seem to have a bug because it tells you to update now inside the dashboard even though you've already updated. Right. So, we are running um version the 2nd of March. As you can see, it's confirmed inside the chat. But when I click it, as you can see right here with the notification, it still tells me to update, right? Which doesn't make sense if we're already updated. So, it didn't break, but it got a little bit buggy. Um sometimes it does break, but yeah, that's basically it. Shackle says, "What's the point in Olama? They just chat. " So with Oama, you can actually run it with open claw and then it doesn't just chat. It's an actual full agent, right? So if you're running with open claw or zero claw or any of these sort of autonomous AI agents, then you get the power of the agent, but you get it for free with OAM running locally. So it's not just a chatbot. Annie, good to see you here. Annie Ma, legend in the SEO space. Thanks for joining. Annie says, "So useful. Thanks for showing us how it's done. AI feels less intimidating now. " Thank you so much. And then Shackle says other than Claude that came with a CLI. Yeah. So I mean like you've got you can access with OAM like tons of different tools, right? It's pretty cool. Um you might not be aware of it, but like if we go to for example like Kim K2. 5 here, you can see that you can run this for free, right? Um if you're sticking within the token limits or just running a local model, you can run these tools with like claw code and codeex and open code and o open claw, right? And these normally are expensive to run with APIs. So you can run them for free using a localized version of for example like Quen 3. 5 and that sort of thing. So that's the point of alarm. It's not just a chatbot.

### Segment 5 (20:00 - 25:00) [20:00]

Click Sanger says, "Hey Julian, do you know a better AI on OAMA than Quen 3 coder? " Yes, you want to check out my tutorials on how to use um Quen 3. 5. Now Quen 3. 5 just came out a few hours ago. Super powerful. You can get different versions of it and it performs at the same sort of levels as like Claude Opus 4. 5. Now, honestly, do I think they're as good as something like uh Claude Opus 4. 5? No. But if you want to run like a powerful model locally, then go with Quen 3. 5. Right? So, basically what you can see here is like they've got models with the new version to match whatever you can run locally, right? So, if you're like on a super powerful computer like a maybe a Mac Studio, then you could run something like the 24 gig or the 17 gig version, right? But if you're on something that's quite slow or doesn't have a lot of power, then you can run the 08b version. That's the most lightweight version. It's only 1 GB, right? So, it can run locally. You can use it offline. Um, and that's quite a powerful model. Let's see what else we got on the questions here. How do you decide the correct stack for a task? So, I mean, for me, it's quite easy to test them out and see what performs, but I think everyone has their own preferences, right? So, like for me, it depends on the task itself. Like for example, if I'm going to write content or do something workrelated, typically I will just use Claude, right? Claude is like the best model um when it comes to most stuff. But if you need something that's like fully autonomous like it's like do this task, you walk away, you come back in a couple hours, then I'll probably do something like perplexity computer. Um but the main thing is it comes with experience. So I've got a lot of videos on like different AIs you can use and I would just look at them and see, okay, which one do I want to use, set up, etc. Right? Um, that's the way that I would try and have a look at it. Any cost effective APIs you would recommend? Yeah, for sure. Like, you know, you've got Miniax, GLM5, and Kim K 2. 5, right? So, these uh these models are like super powerful, but they're nowhere near as expensive as something like Claude. So, I would just go with one of those. They're a lot more efficient. Clicker says, "Thank you, sir. The best teacher on the planet. " Thank you so much. That's great. Um Shackle says who's ah okay yeah so in that situation right I'd have I'd always have the most powerful API as a brain and then it can delegate stuff to for example Miniax communicative 5 or for example GLM5 um as a sub agent right so you can use the cheaper APIs as sub aents and then you have claude as the main brain because that's the smartest AI so inside openclaw you can create sub aents and then just say okay claude make sure you're the brain and then you give the APIs to everything else. Right? That's basically how it works. Um, Urban says, "I'd love you to do a video on null claw. " I've actually already done one. You can check that out inside the AR profit boardroom. So, if you go inside the classroom here and then you go to SOP updates, we have a full video tutorial on null claw. We set it up, we tested it, we created a 30-day road map for it and also 100 prompts on how to use it, right? So, if you want to check that out, that's inside uh here as you can see.

### Segment 6 (25:00 - 30:00) [25:00]

And Urban says, "Your content is awesome. That avatar had me fooled for a few minutes. " Thank you so much. No, I appreciate that. Thank you. Um, we do try and be really transparent about when we use the avatar and that sort of thing, but I appreciate you checking it out. So, thanks. All right. Coport just dropped and it might be the biggest contender to competing with open claw we've ever seen. Now Alibaba, one of the biggest tech companies on the planet, just released their own free AI agent. You can see the announcement right here. So, this is a tweet from AI search and you can see that Tongi Lab actually announced that Coport was available right here. Um, and basically this is running via a free AI agent. It does almost everything Open Claw does. It can run on your computer. It remembers everything. It works on its own 24/7. Now, here's where it gets wild because Copore apparently can actually do things OpenClaw cannot do right now. things that once you see them, you'll wonder how you ever used agents without them. So today, we're going to check it out and just see how it runs. If you want to take a look at it, it's open source. You can see on GitHub right here. This is Copore and this is running from Alibaba, right? This is your personal assistant. Easy to install. You can deploy it. It was fairly easy to set up. What I would say is when you're using this, use something like Claude Code to help you get it set up because it can be a little bit of work unless you're super technical. And you can see an example right here. Now, this is a preview in Chinese, but you can run it in English as you can see. And then to run it, it's just like a oneline install like this. You can also run it using Docker, which is what I did today. And then once you've done that, you can actually set it up with like local models or all sorts of models that you want to use, right? So, here's an example of how it runs. If you go over to the Coport console over here, this is Coport, right? And you can see, for example, we actually created this landing page. So inside the chat, one of the first things I would say is like it did even when I selected English, it started speaking to me in Chinese, right? So you can see an example of it speaking to me in Chinese here. Um, and then what we actually did was we just said like just reply in English and then it switched back to English here. And from there, I actually wanted to run a test to create a website. Right now the website itself is pretty awesome. Like look at this website that it's created and it just built this using the prompt that I'll give you. So, it was a oneshot prompt. It just went off and created the whole page, which is pretty crazy. Like all the FAQs um were created. The colors are pretty nice, very colorful, very interesting. Actually looks better than the original in many ways. And if we go back to Cop, I'll show you how we did this. So, literally you can use it inside the chat here. So, you can say, for example, like build a beautiful um fun landing page that's easy to use and then actually just pasted in the information from the AR profitable boardroom into the chat. Now, from there, it said, "Sure, I'll build you this. Let me create something that hits different. " Blah, blah. And it went off, created it. You can see that it's sort of coding out right there. And it's like done, I've created this. Here's the link to it. So, it's actually coded locally. And here's why it's good, etc. Here's the key sections that it's added. And it's like, if you want any tweaks or anything like that, just let me know. Now, the one thing that I'll say here is like the browser use was not working so well. So what I'd recommend is that if you want it to use browser use properly, just go back into claw code and get it configured properly. But it was pretty easy to set up. And the interesting thing is on the left hand side here, you can see that you can select like channel sessions, chron jobs, and heartbeat just like open claw, right? It really does um it must be like a open- source version that was built around openclaw in the first place. Now if we go to the settings over here, you can see that you can switch between different settings. And I'm actually running this with OAMA, right? So this is running uh we can run it locally with Olama. You can choose which for example um which model you want to use inside

### Segment 7 (30:00 - 35:00) [30:00]

the list. So you can run this completely for free and it's quite easy to set up the API right. So for example we could run this with quen 3. 5. for example GLM 4. 7 and these are both local free models so that number one copore is free to set up because you can just run it from GitHub and then OAMA is free to set up and run locally as well right and then you've got an autonomous AI agent running locally on your laptop just like openclaw and it's just running for free 24/7. Pretty crazy. Now if we go over to the workspace here you can quickly edit your files and that sort of thing. I want to see as well actually. Can we change it to um to dark mode? Right. So, right now it's pretty bright. That's what I would prefer it if it was in dark mode. It doesn't seem to be, but I wonder if you can switch it in some way. I'm sure you could if I played around with it a bit more. And then you can run it with all your main sort of models, right? The main one I would probably focus on if I was you um if you're testing this out is just using the free Olama as you can see over here, right? any bugs or anything like that, just go over to Claude Code and say, "Can you help me fix this? " or "This isn't working," blah blah. That's what I did when I had a couple of problems with it set up. But literally got it set up within about I would say 10 to 15 minutes and it was pretty simple. Now, you can add MCP clients over here, right? Which means you can link this to tools um just by pasting in the MCP details over here with your API keys. And that means that basically you can make this even more powerful by linking it to tools externally and then your autonomous AI agent with copore can go off and start doing stuff. Um the other thing that I would say here is like it doesn't break on me as much as something like open claw and I'll give you an example like a real practical example from today. So for example with openclaw here we've already updated to the latest version but it's still got a notification to update at the top which doesn't make sense if we've already updated right we're already on the latest version. So just something to be aware of there. Now the other thing that I like as well is that you see on this section here you can add multiple chats right which means that you can have parallel agents working on multiple different tasks at the same time. So for example we could say okay code a pomoduro timer over here right then we could go over to this chat and we could say okay code a fun snake game. Right? Something like that. Right? And then [snorts] we can have multiple chats running with the AI agent and just they're all working in parallel together. Right? That's a pretty simple and easy way to just have multiple chats working so that you can have these agents um working in parallel and you can have different stuff running at the same time. Now let's have a look inside the chat here. See what we got. I'm going to come back to these questions actually in a second. So um we focus on this particular update. Now you can see we have multiple sessions running here. So you can switch between sessions. You can set up uh schedule tasks just like you would with openclaw. You've got the heartbeat file which basically is like this will send you updates, right? And so every few hours you can get updates. Um and you can change the interval. So you can say okay update me every hour or something like that. And then you've got sessions and channels, right? So with the channels you can actually link this to Discord. I couldn't see an option for Telegram, but you can link it to Discord, QQ, um a couple other Chinese messaging apps, but probably the main one that most people watching this would use is Discord, but you can use the UI, right? Um and that's basically how it works. So, you just put in the chat here what you want to do. um because it's using no lumber, it's obviously going to be a little bit slower, but you can see it does actually work and it does work directly here. So, that's basically how to use it. It's pretty simple to set up. Would I say it's as good as something like open claw? Probably not. Um it was a little bit buggy when it was using browser control, whereas browser control seems to work fine if you're using something like openclaw. Um but I think either way, it's just interesting to test these alternatives out. And the overall trend here is like things are becoming more autonomous. More agents are being released. Alibaba is obviously taking this seriously because you can see um how it's working. And then you can also see here that we've created this Pomodoro timer, right? Um using the local API from Alarma. So it didn't cost us anything to set that up as well, which is pretty cool. Now, let's just have a little preview of it.

### Segment 8 (35:00 - 40:00) [35:00]

So yeah, that's how the page looks like. Um, pretty simple and easy to set up. And then we've got the timer ready to go here. Right. So, this is the Pomodoro timer we just asked Copore to create, which you can see right here. Um, it's working perfectly. Pretty interesting design. Pretty cool. It's got an alarm as well at the top. And then you can reset the timer as well. And then you actually get like um a streak at the bottom so you can track how many sessions you've done. So, it's all working fine. It seems to create stuff pretty cool. Um the one of the biggest differences I think is like you get the chat options so you can come back to your history and that sort of thing. And then also just seems a lot easier to configure because you can switch between APIs and the settings. So that just makes it a lot easier as you can see. Now, just to recap here, um we do have like a 30-day system for building AI agents with Copore inside the video notes from today inside the AR profit boardroom. Um and you might be wondering, okay, what is Copore? Why does it actually matter? So, it's an open source AI agent built by Alibaba's Tongi Lab team, right? It was fully open sourced on February the 28th and it's got a long-term memory that actually works. It remembers your preferences. No more starting from scratch. Local vector search. It runs locally first. So it runs natively of Llama as you've seen today, which means you can run it for free and it's all like sort of running locally for you. You can connect it to multiple different message channels. And the interesting thing here is you can actually build your own channel plugins, right, with a standardized protocol, which is pretty cool. So the biggest difference I think between these is like if you look at something like Copore, it's really designed and I think this is this sort of coincides with Quen 3. 5 coming out which is local, free, and really powerful from Alibaba. Um I think they've designed this AI agent to work alongside Quen 3. 5. And then you can see here there's a 30-day road map if you want to test this out. We've got day-to-day um prompts for really mastering this if you want to check out and how to use it and that sort of thing. And then we've got 100 prompts as you can see down here uh to test this out. So just to recap, Copore is Alibaba's open source AI agent. It was released end of February. Local first, memory first, model modular architecture. Zero cost because you can run it with alarm. It connects to Discord, iMessage, and a bunch of other messaging apps. In terms of copore versus openclaw, so both are powerful. Copore is more local, right? Whereas open claw is more cloud-based when you're using APIs unless you're running it locally as well with alarm. Um and it's [snorts] just like different philosophies and different UI as well I would say. So thanks so much for watching. If you want to get all the video notes from today, you can get that inside the AR profit boarding. We actually have a full 6-h hour course on openclaw if you want to check that out and another 3-hour course on openclaw right here. And we add new daily updates inside the profit boardroom. As you can see, this is my AI automation community, the AI profitable link in the comments description or go to a profitable. com. And this basically shows you how to save time, scale, and grow with AI. Now, inside the community, we have 2,500 members, so you can ask questions, get help, get support whenever you need it. There's always people online to help. Um, what you can also do, for example, is you can go inside the calendar and get [snorts] video calls, live automation, AI automation coaching, so you can ask questions, meet the community, ask any questions you have. if you ever get stuck, they're great to go on. Inside the map, you can actually zoom into your local area and then meet people in that sort of thing. And then additionally, inside the classroom, you'll get all my best trainings. Now, at this point, you might be saying, "Okay, are people actually getting results with this? " So, we have 133 pages of testimonials at this point. It's basically an endless scroll. And what's cool about this is, as you can see right here, there's lots of people winning with this stuff, right? So, it's it's something where you can go learn AI, get results. I mean, for example, Russell was selling AI avatar videos. Eric landed a new client. Dan got his first customer in two weeks. Like, and there's 133 pages of people getting results like that. So, it's a pretty awesome community and definitely worth checking out. Let's see what questions we've got. — [clears throat]

### Segment 9 (40:00 - 45:00) [40:00]

— So Rowit says, "Hey Julian, I have an old Windows laptop. I'm trying to set up OpenClaw in it. What AI stack do you suggest me to run in it? " So I mean just use the normal stuff, right? So for example, like you can set up Miniax or uh GLM or Kimik 2. 5 if you want a cheap API and then just run it locally in your terminal. You don't need to overthink it. Like it can run um however you want. Faso says, "Updating to 3. 2 now. How long did it take? " Literally took like a couple of minutes for me to upgrade. Copore can use Miniax 2. 5 free. Not as far as I know, but it can run local models. So, it can run Quen 3. 5 free, for example, and it's set up to be able to sort of run the API directly inside the settings with OAM. Is there a possible way to spawn sub agents to delegate tasks on Copore? I think there probably is. I've just not checked that out but I think like for example sub aents which are a feature of open claw you could probably easily set up inside coore as well and then shackle says what about visual and documents for coort yeah so like you can create documents you can create um if you gave it the API details so if you for example said okay here's my API for Google AI studio here's how to set up nanana banana 2 um here's the documentation for it could easily set that up right in the same way that you would set that up with open claw. And then Urban says, "What's the best local LLM for development work on O Lama? " So yeah, I would I mean it's that's a really good question. So I would recommend we've got a bunch of training on it here. What I would recommend is that you check out Quen 3. 5 first because it's got some awesome models that have just been released in there. It's really good for coding, right? Um, you can also, for example, as you've seen today, we use GLM 4. 7 Flash to create this page and also the Pomodoro timer, right? And it coded them out perfectly, right? So, there was no issues. Um, it seems to work really nice. It's done some nice designs like right up there with something board would do. And so, you can easily code out locally with uh GLM 4. 7 Flash as well. It depends on your setup as well, though. That's the main thing, too. some other use cases for this as well like you could use open claw to help you come up with like content ideas or copore as well. Um we actually do have a full sixhour course here on like all the use cases for this sort of stuff. So if you want like examples of recommended skills and what we use for this sort of stuff, you can check it out right here. — Claw X just changed how everyone uses OpenClaw. Most people have no idea exists yet. So right now, OpenClaw is one of the most powerful AI agent systems on the planet. It can work for you 24/7. automate your tasks, talk to apps, [snorts] schedule jobs, all of it. But there's one massive problem. You have to use command line to run it. Um, that's like having a Ferrari, but the only way to drive it is with a manual written in Japanese. So, Claw X fixes that. And this is a brand new free desktop app that puts OpenClaw inside a clean, beautiful um, UI, as you can see right here. And this runs via a desktop app, which is pretty cool. Now, this is a brand new free desktop app that basically helps you use OpenClaw, but you don't need to type commands. You don't need any confusing setup. There's no technical needed. You just install it through a setup wizard and then your AI agent is live. And I actually got OpenClaw to set up directly. So, I'll show you exactly how I got that to work. And we can run through that in a second. So if you go back to open claw which you can see right here

### Segment 10 (45:00 - 50:00) [45:00]

What I literally did and this was really easy to set up. I just got the GitHub details of claw X. So it's a free open source project on GitHub as you can see. I took the URL of GitHub then plugged it into openclaw and said install this for openclaw please. Right. And it kind of creates claw X. And as you can see this is running locally. But the interesting thing about this is it's inside a desktop app. It almost feels like called co-work and that sort of thing. Um, which is pretty powerful. Now, it took a few minutes to generate, but then once it did, it gives me an overview of what it's doing, how it works, etc. Um, it sets it up automatically inside the chat, so you can see. And then it will tell you, okay, this is set up and good to go. Right now, also, for example, asked me, it's it said installed now, as you can see. Um, but it did ask me like, do you want to set this up as an app inside Mac OS? And I was like, yeah, let's do that. Right. And then from here, we actually got the full app as you can see. So, inside the desktop app, we've got Telegram, Discord, WhatsApp. It just makes it way easier to manage and it's a pretty clean UI as well. So, you can see we've got Discord connected here and it gives you kind of like this mission control as well. So, you got like how many connected. Um, you can easily like set up a new channel as well if you want to set up a new channel here. So you can click on Telegram, add the bot token and the allowed user IDs. So it's really easy to connect other apps. You could add WhatsApp and that sort of thing as well. Just scan the QR code. You can also, for example, go to the dashboard here. This kind of like mission control where you can see the gateway, what skills you've set up, everything else. And then you've got all the models you're using over here. Inside the settings, you can change between light and dark. You can change the language, the AI providers that you have. So, it's quite easy to, for example, add uh other APIs like you could even add O Lama and run this for free locally as well, which is pretty cool. And then you can add custom AIs here. But, Anthropic, OpenAI, Google, they're all available. And that means you can just configure the setup of your OpenClaw way faster by doing this. And it's super easy to do. You can also download the latest updates inside this section. You've got the dashboard, channels, skills, schedule tasks, and then the main chat, right? So, you can actually just chat with openclaw um directly inside the chat here. So, if you say, "Hey, are you working? " This will be connected to my open claw. And the cool thing about all of this is you can easily get it to um to open core to set this up yourself, right? Which means you don't even need to like run the terminal commands to set it up as well, which is pretty cool, right? And you can see here it's working directly, which is awesome. So, let's see what we got inside the questions here. Greg says, "Hey, aren't you finding open core so full of bugs? " Absolutely. Yeah, it's very buggy. I mean, that's what you're going to get with like open source projects, right? open source projects, you're always going to get issues where it's like, you know, um where it's, for example, like breaking or struggling or that sort of thing, right? And it happens all the time. So, um I would just be aware that this can happen. There's some issues from time to time. Um but yeah, you could you you're going to get that with like open source projects. It's not like Claude where you know it's a really clean smooth experience. Even like the updates in section here it struggles with. But that's basically how to set up Claw X. Um that's how to run it. You can see here it's running inside the chat. I've shown you how it works etc. Um and you can run it just as a application which is pretty cool. It's also quite easy to create schedule tasks and manage your schedule tasks as you can see. And the other thing here is as well like it's kind of managing all the problems that people have, right? So for example, people struggle with the terminal setup on open claw. So with this, you just have like a one-click installation. The configuration files can be quite difficult, but with claw X, you get like the visual um real-time validation and that sort of thing. And then there's process management, uh multiple AI providers, and it's much easier to like uh add skills and plugins and that sort of thing. So that's basically how to use it, how to set up, how to get it working. Again, I've shown you what it looks like. It's pretty cool. Um, definitely worth checking out. It again, it does feel like almost like the claude co-work

### Segment 11 (50:00 - 53:00) [50:00]

if that makes sense. [snorts] And you can get open claw to just set up inside the chat. Just paste in the GitHub details. So, thanks so much for watching. If you haven't already, check out the AR profit boarding. We've got tons of video notes as you can see right here. Um, we've also got a full framework on how to use this, a 30-day plan on how to set up a bunch of use cases here as well, and also 100 prompts on how you can test this out with claw X and just decide like, you know, do you prefer the desktop app or the gateway or that sort of thing. Um, so that's basically how it works. Now, inside the AR profit boarding as well, we have a full six-hour course on how to master Open Claw with step-by-step guides and that sort of thing. We have another three-hour course here, and we update this every single day with new updates. So, for example, Quen 3. 5 released a new model you can run with OpenClaw this morning. We've already done a video tutorial on full step-by-step guide here. And we actually date the updates. You can see for example, second of March, 1st of March, etc. And we're adding new updates every single day. So, this is all inside the AI profit board link in the comments description or go to the apiprofitboard. com. Inside the community, you can post questions, get help, get support. There's always people online cuz there's 2,500 people inside here. I also do weekly updates. Now, inside the weekly AI updates, what I do is I look at my 60 hours of research each week with AI, and I break down what's useful and what's not, just to save you time and to help you avoid information overload. We also have a daily accountability group. And again, this just helps you avoid information overload so you can stay focused and just really focus on what matters to you most. Right now, inside the classroom, you can get all of our best trainings. So, for example, if you want to go from beginner to expert with AI automation, you can do that here. Plus, learn how to build your first AI agent in under 5 minutes. Additionally, you can get our playbook which shows you how to automate X, newsletters, shorts, avatar videos, and everything else. Inside the SAP updates, you can get daily updates like you can see right here, um, with all the new guides and that sort of thing. And then you can also go inside the map and meet people in your local area. Additionally, you can learn how to get more clients with the agency course. You can watch about the coaching calls. You can learn how to rank number one with AI SEO and also how to grow a YouTube channel with AI based on what's working for me. So, that's all inside the AI profit boardroom link in the comments description or go to the AI profitboarding. com. Shackle says, "Are the tears to claw a better or worse? " Yeah, there's plenty of alternatives I've tested that have just not been as good. Um, we've got loads of guides inside here. Does copore support open router? I think it does as far as I'm aware. I think so. Are there tiers to claw? Uh, we already answered that one. Um, does claw x update open claw automatically? Well, you just set up auto updates inside the gateway of or if you want to automatically update. [snorts] Um, rot says, I want to build an autonomous job application setup. Yes, you can. I I've done that with the way that I did that before. I built a job board with um data button. Data button was the uh one that I used and we have a full tutorial on how I did that right here. Right. So, we have a full tutorial on how data button work right there. So, that's basically how it works. Thanks very much for watching. Appreciate as always. I've answered all the questions. Thanks for being on