NEW Chinese AI Image Generator is INSANE (Qwen-Image-2512)

Image 2512 is out now. This is an update from Quen which is released by Alibaba and we're going to test out today. Now from the results so far, so it's been tested in 10,000 blind rounds on AI arena. Quen image 2512 ranks as the strongest opensource image model and you can actually run it for free on your laptop. We'll come on to that later. I'll show you how to do etc. So if you want to get

access to this, you can go to chat. quen. ai and then you can start using it inside the image generation. You've actually got a bunch of sample prompts here. So for example, like a copy bar holding a sign that says hello world wearing a suit. And then from here you can start using it and it'll actually generate the image. As you can see, one of the things that has surprised me is how quick this generates images compared to for example Nano Banana or Chat GBT. And as we run this, you can see an example of the image generated like this. Now what you

can also do is you can click on edit and you can actually edit the image. So for example, you can say make the suit bright green and we could test out actually editing the image. So this was the original and then you can see it's generated the image exactly how we wanted. It actually looks pretty good to be fair. And then over here we've changed the suit to green. So you can use it as an image editor as well which is pretty awesome. And then over here what you can also do is you can click on create video and then from here you can create video or you can download the image. So if you download the image that's saved now and if you want to turn it into a video you can just sign up with Quinn and then turn it into a video. Let's test this out. And bear in

mind you can use this without signing in. So I've just had to sign in now but before we didn't. And if we go back to the image prompts like this. Let's try another one. We're going to generate this image here. And now we've got the image itself. All right. Now once you generate the image if you click on create video. So actually switch to video generation. And then you say okay turn this into an image. And that will take a bit longer to generate. But you can see that'll generate an image for you as well. Bear in mind like I'm not on a paid subscription or this is on the free plan that we're actually using. And you can use Quinn for many other things as well if you've never tested it out. So if we go to settings here you can see my account is just on the normal free one. might not pay to use this or anything. So that will take a little while. But actually what I like is the progress bar here. Normally you don't get that inside for example Gemini. You just have to wait for V3 to generate the video. Whereas with Quinn, you actually get a progress bar. And then the images themselves, not bad at all. All right. So let's go to a new chat here. And you

might be wondering, okay, like how to use this, where are all the resources, etc. Where are the links? So, if you go to the AI success lab link in the comments description, that will actually connect you with 42,400 members and an amazing community, but also if you go to the classroom here and then you go to December 2025, you will find the latest video notes plus of resources on how to use this stuff. All right, you can get all the links to everything here. You can also use the API via Alibaba code as you can see. I think it's embedded inside foul now as well if you want to generate images that way too. You see the price per generation on the API is pretty cheap to be fair. And then if you go to the GitHub and the hugging face and the blog, you can see some examples breaking it down right here. Right. So coin image 2512 finer details great realism blah blah. And then you can see the GitHub here. So it is open source and you can test it out directly inside the chat or you can test out on huggy face as well. So you can actually launch it. Bear in mind if for example I'm on a M3, a Mac M3 is probably not going to run. So just bear that in mind if you want to run it locally. It's got to be pretty powerful laptop I think. But you can see the images leaderboards as well here. So for example, if we have a look here from the leaderboards, Gemini, Google's image gen are right at the top here. Crim is number three. And then we've got Quen number four 2512 by Alibaba. And you can see a win rate of 40% here. So it's scored pretty well there. And some of the key points to note with this image update are enhanced realism, finer natural detail, and improved text rendering as well. I'll test it for My hopes are not too high for that, but we'll test out. So, here's an example. So, this is the old version. This is the new version. And this one looks a lot more realistic than that one. That looks more like UGC on the image. Here's another example. Another one. Let's have a look here. Yeah, they do look slightly more realistic, don't they? And then here's an example of a waterfall. And you can see like this is more cartoonish. This is a lot more photorealistic, a lot more detail inside. The same for this. The images now much better than they used to be. And I have tested quen image in the past for using this stuff and it just wasn't that great. Let's test out for

text now as well. So I'm just going to take this prompt. We'll test it out inside the chat. Let's try out now. Plug in the prompt. So we're going to try and generate an image. You can actually change the dimensions. So I know inside the prompt here it says change the dimensions. That won't work. We need to change the dimensions inside the drop down there. JM says Quen question mark. Yes. And then A says, "What what's up you cheeky geyser? " Awesome, man. Good to see you. All right, so let's have a look at this. I mean, that image is pretty cool actually. I like the image that would actually fit pretty nicely about a profit board images. So, if we go to the AR Pro boardroom here, classroom, you can see the image styles that we generate and that one would actually fit pretty nice like this branding similar. The text is rendered perfectly inside the image. So, air proper boardroom and then 1,900 members, daily AR updates, live coaching. Pretty good image to be fair, like pretty nice. It's obviously you would want to make sure that you choose the right dimensions. So, don't put the dimensions inside the prompt. Put it inside the dimensions over here. But yeah, it's

working nicely. It says, can I ask what's the difference between SEO lead circle mastermind? Yes, honestly, I would recommend that you check out the air profit boardroom because if you actually sign up to the air profit boardroom, you get the SEO elite circle for free as a bonus. Yeah, I would do it that way. Just check it out. But the SEO lead circle is a community for learning SEO and the AI profit boardroom is more a community for just learning AI and all our best systems and that sort of thing. So two different communities right there. So that's basically it for using Quen short and sweet. We've tested out we've tried a few different images. Let's test something else out now. And I will do a deeper test here. But I want to test a couple of things here. So number one is speed. And number two is the quality. All right. So, we're going to plug this prompt into Quen images like so. We're also going to do the same thing inside Nano Banana, which is I think by far the best option for creating images. And then we'll test it inside chat GPT images as well. Create image. There we go. All right. And we'll test these three out. Right. So, you can see that Quen generates the image super fast, but the image itself just feel it still feels like AI, doesn't it? When you look at that, you think it's not photo realistic, right? If you click on edit here and we say make this photo realistic and just see if it can improve it, especially EGC style. We'll see if we can improve the pro. But that's the only thing that I would say here is like when you're creating these images, and you'll see why in a minute. If you look at Gemini versus Quen, Gemini's images are way more realistic, right? They just feel a lot more like actual humans. And if we have a look here, you can see Chachi 5 is no, we're not going to do that for you, mate. help you out there. So, it's okay, mate. That's how it goes. chat cheap lets us down as usual. It's one of those and yeah that's basically how it goes. So the massive mind for three a month that's something different. Yeah. So we actually that was an old off we don't have but yeah so just to recap chat

cheaper Quen image 2512 it is pretty decent. It's great for editing images. I think that's honestly the strongest point I've seen today is not as competitive as the images from Gemini but it does outperform chat sheep so far. I might do a video on how to test it properly and do some more in-depth tests, but that's it. And it's free. It's open source. You can get it on Hogface. GitHub. It has just dropped less than 24 hours ago. And pretty impressive so far. It is out of all the open- source image models, it is the number one leader as well. So, that's something to bear in mind. And you can see a clear difference. The other thing I like about it is the text rendering. So, it is pretty good for that. If we have a look and we go back here, you can see that the text rendering for this particular image is pretty it did it perfectly, which is great. So, thanks so much for watching. If you want to get all of the video resources from today, feel free to get that inside the AI success lab. This is an amazing community with 42,400 members. And you can see if we go into the classroom here, you can get all the video notes from today. So if you want to grab all the video notes, all the resources, if you want to grab some example prompts, I've actually plugged in 100 prompt inside this list. So you can just grab them for free. So you can grab my 100 AI image prompts. On top of that, you can also get a 30day plan as you can see right here. The other thing as well, this is interesting. So some people used to say AI can't do text in images properly anymore. And the old way was like misspelled words, gibberish letters, unusable for business. But the new reality is like Quen image 2512 and particularly Nana Banana Pro were built from the ground up to render text. So it would be pretty good for like infographics, posters, presentations, that sort of thing as well. So pretty good. I think the new empowering belief here is like AI image models can now handle text better than most humans can in Photoshop. The other cool thing about this is some people think you need technical skills to use this stuff, but you can just go straight onto the chat and then you can just create images directly there. So thanks so much for watching. If you haven't already, check out the AI profit boardroom link in the comments description. It's an amazing community where we have 1,900 serious AI builders focused on getting results with AI. Inside there, people ask their questions, post their wins. I personally answer the questions inside the community and also you can see that everyone helps each other out. It's a very active community with some of the best AI builders I know on the planet inside this community. On top of that, if you want to speak to someone live on a call, you can actually do that four times a week inside the calendar here. So, you can jump on the live coaching calls, get help, get support whenever you need it. Also, meet the community live on a Zoom call. And then also, you get our best courses. For example, if you want to learn how to master AI in just five weeks, and go from complete beginner to expert, you can do that in five weeks. Plus, we'll give you a whole framework for building your first AI agent in under five minutes. Top of that, you also get our best playbooks on how to do all this stuff. So, for example, like X automations, shorts, how to automate AI avatar videos, and you'll get my best trainings on N10, how to get clients for your agency, our live coaching call recordings, AI SEO automations, and YouTube AI. So, that's all inside the AI profit board. Feel free to get that link in the comments description.