Magnific AI Masterclass: Build INSANE AI Workflows (Freepik Spaces)

追加: 2026-05-18

[00:00:00]In this video, I will show you how to start using AI web flow builders to create insane visuals using AI. And to do this, we're going to use my favorite platform that is Magnific Spaces, which was called Freepix Spaces not too long ago. And I think right now it is the perfect time to start learning how to use these workflow builders because all the most successful AI creators that I have talked to are already using these every day. That said, in this video, we're going to start from a complete scratch. So, I will walk you through how to use this platform, what are all these different nodes all about, why you should care, what are my favorite workflows, and I will also build some workflows with you in the video. So, this is going to be a complete tutorial from zero to hero. And by the way, everything that I show you or build in this video, I have left a link to directly in the description. And also, you will find a really nice discount to Magnificaces there as well. But that said, let's get started with the video.

[00:00:54]And now if you have never used AI workflow builders before, you may have encountered some of those workflows and you have been like, "Wait a second, that is way too complicated. That looks almost like some programming language.

[00:01:06]So I don't really want to learn that."

[00:01:08]And actually, I had a very similar impression about those tools before I started to use those. But once I understood that they're actually super easy and intuitive to use, and once I understood the power of them, I actually fell in love. So much so that I basically do everything inside AI workflow builders these days. And before I show you how to start using magnific.com, let me actually very briefly show you how these work and how they differ from these regular image generators. So as the first example, let us open up the good old chaty BT and let's just tell it something like create an image of a pair of shoes and then let's hit send. And here is the picture.

[00:01:50]Now let's do this exact same thing using an AI workflow builder that is magnific.com spaces. So I will just sign up. I'll go to the spaces tab on the left hand side. Click on new space. And then I'll just need to rightclick over here inside the canvas. And I can for example click on the text prompt. Then I can just type in create an image of a pair of shoes. Then I'll put it aside. I will drop a connection to an image generator node and I will hit run. Now, this is the exact same thing what we just did over here inside GPT as to what we did here in the magnific.com workflow builder. And there we go. Here is the result. But now, why on earth would you use a workflow builder when you can do the job much easier, for example, on CHPT or for instance here inside Hicksfield AI? Well, first of all, if you use Chad GPT to create your images, you're only using the Chat GPT image generator model. So, that is only one way to create a photo. And you already know that there are all these models like Nana Banana 2, Nana Banana Pro, you have Google's models, you have all these different AI image generator models, and you can only run one of them here inside ChachiPT. But here inside magnific.com, you can actually choose any one of these AI image generator models. So if you want to use the flux models, you have those over there. You have all these Google models over there. You also have the GPT image 2 model with which we just created this pair of shoe inside chipt.

[00:03:24]So you can even run this as a chipt conversational and you have the gro model and everything listed here under this one subscription. So now if you for example find a use case where you're trying to create an image and you just can't nail it. The AI just doesn't produce a good result just go over here and choose another model and you'll probably get a much better result. And that is just something you cannot do over here in JPT. So that's one side of the equation. But if we think about a platform like hixbuild.com or open art, you can actually do the exact same thing over here. So if we tell to create an image of a cat, we can choose any one of these AI image generator models right now. So why wouldn't you just use a simple interface like this that allows you to access all these image generated models under one subscription? Well, that is because these AI workflow builders are not there just for you to create this kind of simple workflows. In fact, I would say that if you only want to create an image or two or if you want to create a video or two, I would much rather use a platform like Hicksfield or Open Art. And actually, I recommend that a lot of times. But if you want to do anything more complicated than that, I definitely recommend using a workflow builder like magnific.com. And now I will show you a concrete example. And by the way, don't worry if you don't understand anything that is going on over here. You're not even supposed to if you have never used a tool like this before. I will show you exactly what this does and what all these node types are in just a second. But for now, I'm just demonstrating why this kind of a workflow builder is such a powerful tool and why it clearly wins platforms like Hicksfield. So here I have created an AI headshot generator workflow. So I'm inserting an image of myself and then I'm turning that into a bunch of professionallook head shot like this.

[00:05:17]And then I have even selected my favorite one. Now, this is still something you could definitely do inside Hicksfield, but there are a couple of elements that make this workflow much, much easier to run here inside magnific.com. Well, first and foremost, if I now wanted to, I could literally just replace this input picture by right-clicking and clicking on content and replace creation. In fact, that is what I just did. So, I found this picture of a man from Unsplash. I just dropped the image in and I hit run on the entire workflow and it created me a bunch of good looking selfies of this person. And then I even chose my favorite one of him. So that just showcases how this is almost like a small AI application where you can just change the input very neatly with just a click of a button and then you will run the entire process on that picture to get all these different outputs. And you can only imagine that doing something like this is simply not possible here inside Higsfield AI or for example here inside Chachi PT and a small technical correction. So in fact in Higsfield these days you have the canvas application which is a similar workflow builder like magnific.com. So technically yes you can do the same thing over there as well. But let's not pay attention to the details because I think the best platform right now for this is still magnific. So we'll be focusing on that in this video. So if you want to create anything at scale and if you want to build repeatable pipelines and workflows with which can actually make content faster and more efficiently than before, you have to use a tool like magnific.com. And it's not just that because here everything happens inside this one smoothly flowing pipeline. So even if you have never used a tool like this before, you can still have a hunch of what is going on over there. So I'm inputting an image. I have a list of scenes and then all of a sudden I am in all of those scenes. So this is actually something that is pretty intuitive to follow. So this is what makes these AI workflow builders such a powerful weapon. And this is just one example. I'm going to show you a whole bunch of more even more impressive and useful examples than this one. But before we do that, we need to jump back to the basics. So my goal is that after watching this tutorial, you will easily understand what these types of workflows do and what all of the favorite workflows that I have linked to in the description also do. So that is why magnific.com is such a powerful tool and that is why I truly believe that these AI workflow builders are the future of content creation. But now let me show you how to start using magnific.com from a complete scratch. Even if you have never used a tool like this before and even if you did not just watch the intro. First things first, what you need to do is you need to go to the link in the description, you will find a nice discount over there for this annual premium and premium plus plans. And then once you have signed up to magnific.com, you will see this kind of a dashboard.

[00:08:27]Then what you need to do next is click on spaces over here and this will open up your spaces dashboard. So this is where all your workflows will land once you start to create those. Then just click on new space and here is your empty canvas or empty space. And now before we actually dive deeper into the workflows and node types, let me actually show you how to move around over here. So just to add something over here, let's add a text node in. And now what you can do is if you hold control or command while you scroll with your mouse, you can zoom in and out neatly like this. Or if you want to scroll up or down, just use the mouse scroller.

[00:09:06]And then if you want to move left and right, you need to use the trackpad with your two fingers. And when you're using trackpad, you can also zoom in and out with your fingers when you do a movement like this neatly. So you can actually control this whole thing with your trackpad if you want to. And then one thing that always happens to me is that I accidentally scroll too far and I don't find my workflow anymore. So if that happens, you can actually just hit recenter and it will lead you back directly over here into your workflow.

[00:09:37]Oh, and something that is also really good to understand right off the get- go is that one of the powers of a platform like this is that you can share your workflows easily with for example your audience or for instance with your teammates and they can join and start to work on the same project. And if you want to do that, let's say you have created a nice image generator workflow over here and you want to share this, just click on share over here. And then you can choose the access level. So for instance, we will do can view since I'm sharing this with you and I don't want anyone of you to accidentally mess up the workflow for the future viewers. But if you for example want to invite your teammates, you can just do canedit. Then just copy this link over there and send it to your friends and they can join this platform right as you see it over here and they can start to do work over here as well. So that is the very basic introduction on how to physically use magnific.com. And speaking of using it, this is an online tool. So you don't need to install anything. Everything lives here in magnific.com. So all you literally need is just a web browser and an account and you can start to build this beautiful workflows. But now let's start to take steps towards building our first scalable content workflow. And to do that, you will need to understand what these most important node types are and what they do. And what would be a better way than me showcasing each and every single one of these with a clearcut explanation with a good and concrete example. So let's do exactly that. So now I have started a new workspace over here or a workflow and I will right click over anywhere in this canvas and I will click text. As you can probably imagine text node allows you to type in text into this node. So the text node is very self-explanatory. So let us jump into the image generator node. And yes that is intuitive as well but there are some things you might need to pay attention to. So I recommend watching this whole thing without skipping. So, first let's type in something like create an image of a dog swimming in a lake in Finland. Let's do something like that. And then let's zoom out a bit and let's drop a connection to an image generator node. And now, as I suggested before, this workflow builder also acts as an all-in-one AI platform. So, instead of having to have all these different subscriptions to all these different AI image generator models, you can just choose the model over here. For example, we can do the GPT image 2 model right over there, which is the latest state-of-the-art best AI image generator out there right now. Then what you can also do here is you can choose the number of creations. So you can for example create 10 images if you want to, but we will keep that at one just for the sake of demonstration. Then here you can also choose the aspect ratio. So for instance, if you're posting this on a social media account, you can do 9 to 16 and that will do a bit of a taller picture for you. Then here you also get to choose the resolution. For example, I think 2K is pretty decent. And then you can also do the quality between low, medium, and high. And let us hit generate. And while this picture is creating, you can actually create another picture next to that one. So let's do another one while this previous one is creating because I have an important piece of news to share. So, one of the things that I absolutely love about magnific.com is that you can actually run these nano banana models in an unlimited fashion, which basically means that you are not spending any credits whatsoever. So, as long as you are on a monthly plan, that is enough.

[00:13:10]You can run this as many times as you want to. And I think they are almost as good as the GPT image 2 model. In fact, when it comes to editing pictures, so if you do image to image stuff, it is actually arguably even better than the GPT image to model because of those weird artifacts that you see in those pictures. And basically, I use net banana in nearly all of my workflows.

[00:13:33]So, this is a gamecher that you can do so for free as many times as you need to. But now, here is the picture we just created. That looks pretty decent. So now what you can also do is you can not only create images from text but you can also create images from other images. So let us drop another connection from this picture which just creates it into another image generator and this time let's do the nano banana pro because I think it is such a good tool for editing images. Let's do a taller picture as well. Let's do the 2K resolution so that we're not spending any credits whatsoever. And then you have two options. So you can either describe the changes you want to make directly over here. For example, let's see something like turn this into an image taken in summer. We can do it like this. And we can hit generate. Or since we're using a workflow builder, we can also drop another text node on top of this picture and do the exact same piece of instruction over there. And then just connect that into the text port of this image generator. And then we can just hit run. And there we have it. Here's the exact same picture, but now it looks like it is summer, which makes a bit more sense in a context like this. Now, since we're talking about an AI workflow builder, the power in this is that you can run this whole thing from a complete scratch with a slightly different input.

[00:14:56]So, for instance, let's say that we want to create an image of a black dog swimming in Finland. So, I will just type it in over here. Then just come over here where it has the run button and click on it. And you can either use or run this node only if you want to just create this picture or you can either also choose run from here which allows you to first run this node and then based on this result run the second one. And there we have it. And now if you want to download your pictures you can choose this whole thing and then you can hit download and it allows you to get all these pictures created in this workflow or you can just rightclick on any picture and download that separately. And now something that you can also do when you're building these workflows. And this applies to any workflow, not just text nodes or image nodes, but any workflow you can ever create is that you can group this entire thing by scrolling the mouse over there and dragging it like that. And then right click and click on group. And now you can give a name for your group. So for example, let's do dog image generator. Obviously on a small workflow like this, this makes absolutely no sense because you can clearly see what's going on. But if you have, let's say, 50 different nodes and you have different types of generators or modules. So for example, five nodes are responsible for creating the scenes. One of them is responsible for generating the lighting.

[00:16:15]Some of them are responsible for the angles and such. it makes much more sense to group those in these logical groups so that it's easier for you to follow if you come back later or especially if you're sharing this with your teammates or audience. But now that is how you use the text and image generator nodes in magnific.com. And now the logical next step is to move on from image generators to video generators.

[00:16:40]And by the way, you will find this workflow that I will be building throughout this tutorial as a link in the description. So if you just want to see the results, the prompts, if you just want to clone it and practice it by yourself, feel free to do so over there.

[00:16:54]But now let's move on to the video generator side of things. So very similar to how you can create images, you can also create videos. So once again, we're going to drag and drop a text node into the view. And this time, let's say something like create a video of a man walking in New York. And with this prompt at hand, we can connect that into a video generator node. And right now, the best AI video generator model available is called Cedance 2.0. Well, that might change quickly because Google is just about to release the VO4 model, but for now, let's run with that. And also here, you get to choose the aspect ratio. So, once again, if we're posting this on a social media site, we can do 9 to 16 aspect ratio. Then we can choose a bit of a longer duration if we want to, but for now let's just do 5 seconds. And here you can also choose the quality between 480p all the way to 1,80p.

[00:17:53]And for this one, let's do 480p because I think that's usually already pretty decent looking footage, especially on a demonstrative purpose like this. And now all we need to do is just click on run and it will start to create the video.

[00:18:05]And there we go. Here is the video of this man walking in the busy streets of New York.

[00:18:16]And as you can also hear, we have the sound enabled over there. So, it actually created some ambient background noise as well as some music to this shot. Although, in this particular case, I am definitely not a huge fan of this music. I would have just loved to hear some ambient background noise and maybe some footsteps and nothing more. But that's besides the point because this just shows you how you can create videos here inside magnific. And now that is just one way in which you can create videos because what you can also do is you can start from an image. So let us actually clone this small workflow and I will copy paste it right over there and let's tell it to create an image of a man walking in New York and let's actually replace this with an image generator like this. use Net Banana Pro this time around. And then let's just hit run. And here is our picture. And now what we can do next is we can connect this picture into a video generator node like this. And then before we hit send, we obviously need to give it some instructions. So let's just say a man walking in New York. And let's just hit generate. And there we go. Here is our fantastic looking video. But now this is pretty basic stuff. So let me actually show you something pretty cool that you can also do with these video generators and that is the start image and end image technique. So let us actually first start by creating another picture over here of the same man in the same spot but riding a bus. So let us drop a connection to another image generator node. I will drag it over here below this original picture. Then let's tell the AI something like this. create an image of this man sitting in a bus just a moment after this picture is taken. Make him look at the distance. So I want to basically create a scene where this man is first walking and then he hops on a bus and then he's walking or watching out the window. So let us hit run and let's see what kind of result we get. And there we go. Here is the picture. And now what I will do is I will create a bit of a longer video where the man is first walking and then he takes the bus and sits down and starts to look at a distance like this.

[00:20:29]And to do this, what we're going to do is we're going to start from this image.

[00:20:33]So we want the video start from this exact frame. So let us drop a connection from this image into a video generator.

[00:20:41]But now, as you can see, this connection by default goes to the references port of the video generator. So that is actually not what we want. So we're going to click over here to cut the connection. And what we'll do instead is we're going to connect this picture into the start image port of the video generator. So this basically forces the video to start from this exact frame. No matter what is going to happen in the video and how much the AI might or might not hallucinate, it will always start from this exact frame. So that will give us full control. And then also very similarly to what we just did, we're also going to drag a connecting line from this end frame and we're going to drop it into the end image port of the video generator node. And then we'll need to just describe what we want to happen in between. So let's say something like a man walks into a bus, takes a seat, and starts to look at a distance. And now since we are going to do a bit of a longer clip, let us do for example, say 15 seconds. I think that is pretty good. And let's also generate sound. And let this time let's do 780p resolution. So we're going for a bit of a better quality. And then we'll just need to hit run and wait for a couple of minutes for the AI to create the video.

[00:21:59]And now there we go. Here is the video that we get back from this one.

[00:22:18]Now, obviously, this is a total catastrophe because it looks like the bus is reversing and swerving between the lanes horrendously before luckily avoiding all the crashes and picking this guy up. So, this is a total mess.

[00:22:34]But to be fair, we didn't pay too much attention to the prompt. And also the AI just sometimes tends to do stuff like this. So this is actually a perfect example that shows that you need to hit run a couple of times, maybe create the same shot like four times, maybe even eight times to get the ideal one. And also you can fix the prompt while doing that as well. So even if you had a more detailed prompt, you would still sometimes encounter issues like this. So that is just something is good to keep in mind. It doesn't matter if you're using Cance 2.0, Vo 3.0, So whatever you might be using, it is always a bit of a risk when using AI that the result turns out not exactly what you wanted or something that is frustratingly close to being actually something worth using, but then it's not quite there just yet.

[00:23:21]So that is the basics on how you use AI video generators inside Magnific.

[00:23:27]So you can either create one from text or you can use an image to create a video or then you can also use the start frame end frame strategy where you have two images and the video starts from the image number one and ends to the image number two and then also a bit of a more futuristic approach is to if you get a bad result to edit the video with a video generator by connecting the video into for example Ceden 2.0 video generator node. But now before we move on to another super powerful node type here inside magnific spaces, let me actually show you something that applies to both video and image generators. And this is especially useful if you have a lot of reference images. For example, if you're adding six objects into one image or six people or whatever it could be or if you have a superdetailed workflow, this is an absolute mustnote trick. Now I have added a picture of me and a picture of this random person I found from Unsplash. And now what we can do is we're going to rename these pictures. So for example, let's call this one dude one. And let's call this one dude 2. And now what we can do is I will add an image generator over here on the right hand side. and let us connect both of these images into the references section of this image generator. And now what I want to do is I want to perform a face swap. So I want my face that is the dude number two face appear in this picture in the dude number one. So now if I give it some instructions I can use the add symbol to reference to either one of these pictures. So for example, let's say something like this. Replace the face in and then add symbol dude number one with the face of then the ad symbol and dude number two. And then let's just hit run. So now this avoids us to having to deal with the confusion that the AI might have when you have a similar looking picture of somebody. So, it wouldn't help to say like replace the face of the dude number one with dude number two if you didn't have these namings in place because the AI wouldn't have any idea as to which one of these pictures is you and which one one of these pictures is the other dude or whatever it might be. And now obviously this result looks absolutely freaking terrible because we didn't take the lighting into account but this proves the point. So it took my face from this image number two or dude two image into the dude number one image. So it didn't get confused as it would get if you just told it to replace one face with another or something like that. So you need to be specific with your prompts. This saves you some time and headaches and hassle. So, if you have multiple objects that you're combining in an image or if you're wanting to do stuff like this that I just did, make sure to rename your images by doubleclicking the name and giving it a descriptive name and then just using that with the at symbol inside the prompt. Then, let's move on to another super important node type here inside magnific.com, and that is the assistant node. If you're having trouble understanding what this does and remembering all of the stuff that I say in this video, this is essentially a chat GPT node. If we go back here in the chat GPT and we say something like create a poem of a cat and we hit run, we will get a result like this. And now what we can do is we can do the exact same thing here inside Magnific. So, we can just tell the AI to create a poem of a cat. And then we can run one of these chat GPT models. For example, GPT5 mini.

[00:27:09]And then if we hit run, the AI starts to process the request. And in just a couple of seconds, you're going to see a poem of a cat. And there we go. It took less than 15 seconds. And this is what I really love about magnific.com because you have everything over here. So you don't really need to have a chachi subscription because you can have similar conversations over here directly. So you don't need to jump between windows or subscriptions at any time. And this kind of a node can be very powerful if you don't really know what kind of a prompt you should write.

[00:27:43]So let's for example say that we want to create a cinematic picture of let's say a young student or something like that.

[00:27:51]So, what I can do here is I can tell the AI, I want to create a cinematic picture of a 30-year-old Finnish student working on an allnighter in an empty classroom.

[00:27:59]And then I can follow that up by saying, give me a detailed prompt that I can enter into an AI image generator to get a good result. And now I can just hit run. Then we'll need to wait for a second for the AI to produce the result.

[00:28:11]And there we go. Here is a detailed prompt. And now I can directly connect this to an image generator node. And we can choose, for example, the GPT image to model. Let's do a cinematic aspect ratio. And let's just hit run. And there we have it. Here is our picture. But after all, this is some basic stuff because you could have done the exact same thing over here inside a Tachib conversation. But now it all happens here inside one canvas, which is super super powerful. But now let us move on to the most important node type in the entire magnific.com platform and that is the list node. And this seems harmless.

[00:28:52]This seems like a boring list node where you just list things but this is surprisingly practical and useful once you learn how to use it. So a list node allows us to list things inside one node instead of having to create tens of difference of nodes. So just an example, let's create a list of animals. And to add these elements, all you need to do is just click on add element and type in something. And then you can just save.

[00:29:17]And if you're not happy with something, you can hit the delete button. And you can also reorder these elements inside your list. So that is the basic stuff.

[00:29:25]But now what we can do here is because the AI is wired in a clever way. If you connect a list node into an image generator node, the AI is automatically going to take every single one of these elements from the list one by one and it's going to feed it through the image generator node. So now if we actually hit run, we're going to create five images with one click of a button. And there's nothing specific about the number five. You could have 500 images or descriptions of images inside this one list and you would be able to turn those into images with one click of a button. And better yet, there's no queuing mechanism over here. So the AI is actually starting to process these in parallel. So even if you created like 100 pictures, you wouldn't still need to wait for more than a couple of minutes in total. So there it is. Here is our list of images. And also it is going to stack those up here on the left hand side of the list node as well. And now since we got a result as a list node back, what we could do here is we can connect this list into for example a video generator. And we could tell the AI to turn every single one of these pictures into a video based on a specific prompt. So that is a power of a list node at its core. And now this is surprisingly effective if you want to avoid repetition. So if you did not have a list node, what you'd need to do is you would need to create a separate node for each of these animals. And imagine if you have 50 different objects or descriptions, environments, whatever it might be. That would be a total catastrophe if you had to add all these nodes on top of one another here and do the same repetitive task of connecting each and every single one of those into a video generator. And then if you need to do a small adjustment, you will have to go through all of these different nodes instead of having to just use this one simple node. But that might not look too impressive just yet. So I will actually show you a couple of examples and this will act act as a segue to those favorite workflows that I have built because in pretty much every workflow that I have used I have used those assistant and list nodes to my advantage. Now so let's create another example below this one over here. And this one's going to be using the assistant node in combination with a list node. So let's do something like this. List me 10 most popular animals in the world. Just give me the animal names, nothing else, please. And now before I hit run, I will actually change this setting from as text to as list. So now it will create us a list of the most popular animals on the planet. And there we go. And now we can once again use this as a list and connect that to an image generator. And we can create pictures for every single one of these animals listed over here. And one by one, the AI goes through the list of the animals and creates an image of those and puts those into this other list.

[00:32:31]Now, this is much faster than what you need to do if you, for example, used ChachiPT to create these pictures because it allows you to only create one at a time or maybe three at a time, but imagine you had to do 100 of these simultaneously. That would take forever.

[00:32:45]But then also something that's even more important is that now that you have these list nodes, you can avoid so much manual work because imagine you are working on a project and let's say that somebody tells you to create these images of these animals that we just did. But they forgot to mention that they would like you to create those images in winter. So each of these animals should be placed in a winter environment. What you'd need to do is you need to go through every single one of these elements inside the list and change the prompt to be for example dog in winter, cat in winter, horse in winter, and that would be absolutely mindless because that is just a pure waste of time, especially if their list grows longer. Now what we can do instead is we can take this listener we just created and then we will drop a connection back into an AI assistant node and then we can tell the AI to turn every single one of these animals in this list into a prompt with which you can create the image in winter time. So I'm going to write something like take every animal from the list and turn it into a short prompt where we tell the AI to create an image of that specific animal during winter time. And once again, we're going to export this as a list. And we'll make some space over here before the list is completed. And there we go. Here is our list of updated animals in their winter environment. And now, since this is a list, we can once again take it and connect it into an image generator node. And we can just hit run. And the AI should create a similar list to what it did before. But now, each animal should be in their natural habitat during the winter time.

[00:34:22]That is beautiful because what we now did is we saved a ton of time. Imagine you had 500 images in your product catalog and then your client tells that, "Hey, I don't like the green background.

[00:34:35]I want it to be pink." Well, well, well.

[00:34:38]If you have 500 different nodes, it's going to take you hours and hours of manual work to make sure that everything is updated accordingly. But here you can literally just take your list node, drive it through an AI assistant node to create another list where you have your edited instructions on each element inside this list and then you can take that and for example turn that into a set of images or video. So this is an absolute huge timesaver if you know how to use the assistant node as well as the list node inside magnific.com. And now you actually understand the most important node types here in magnific.com. That is the text node, image generator, video generator, assistant node, and the list node. These are basically the only ones I ever used.

[00:35:24]And all of the rest here are either some utility nodes or collaboration nodes or something you might stumble across every now and then. For example, in a video combiner node, you can combine multiple 15-second video clips. or a video upscaler node allows you to enhance your video quality if you end up with lower resolution. So this is the kind of basic stuff you have over here and these are mostly for utility and random use cases and I actually have made a separate guide where I show even how these more specific and more niche use case nodes work. So if you want to check that one out, make sure to check the link in the description. It's another YouTube video on my channel. So let me actually dive into some concrete example workflows that I have built here inside magnific.com to give you an idea of what is actually possible. And now because I showed you how this platform works and how the important node types work, you actually now have a pretty good chance of understanding pretty much everything that all these different nodes or these workflows actually do. So, one of the most powerful AI use cases right now is creating these kind of UGC or usergenerated content style advertisements because these are short and the AI videos are always short and these are also something that are non-polished and look non-professional which is also something that the AI is really really good at. And here I have created an UGC style advertisement for this eim product. And basically what I have just done is I have described my influencer with a text description node.

[00:36:56]So there's nothing special going on over there. I have just renamed this node influencer description number two. Then I have connected that into an image generator node and it has given me this kind of a picture. And then I have dropped an image of a screenshot of the app that I'm promoting. In fact, I have generated that picture with the GP image model as well. So this is not even uploaded by me. This is just another picture I generated. And then I have connected that into an image generator in combination with this influencer. And I have told the AI to create an image of the influencer. As you can see, I have used the hashtag or the at symbol to tag the influencer along as well as the app picture that we have over here along into the conversation. So, the AI knows exactly what to do. And it has ended up with a picture like this where we have the person in this first picture holding the phone with this second picture. And then all I have done is I have given it some basic script. So, I'm telling the AI to make it say something like, "By the way, one thing I can't help but recommend is the SY EIM app." And then we have a video result like this. And now the only reason why this workflow is super long is because I have also created another advertisement using the exact same strategy. And then I have also made one for this skin care product. I have actually left links to all these if you want to understand these by yourself better. Make sure to check the link. So you can just come here and start to mess around with it if you clone it for yourself for example.

[00:38:23]So that is the workflow number one. But then there is another one that I have to show you and it is this one that we already checked in the intro because now you can actually understand what is going on over here in this AI headshot generator workflow. So once again I'm just turning myself into a professional headshot like this and not just one but a full-on list of all these headshots so that I can use those on my LinkedIn profile picture. So what I have done here is I have dropped an image and then I have used the assistant node and by the way this is now called scenes because I named it scenes but this is just a good old assistant node that I have dropped into the conversation and I'm telling it to create a list of 20 most common environments for a professional headshot and describe those shortly. For example, creative studio light and as a result I get something like this. And now since this is a list, I have exported it as a list because what I have done next is I have connected that list of these scenes into an image generator node and I have also connected my picture into the same image generator node. So basically I'm telling the AI that here is the picture and here is the environment where I want to place this person in this picture and then one by one the AI converts my face into these environments neatly like these pictures over here and then that generates us another list and then just to showcase what is the best picture I have downloaded one of these. It was this one or this one actually I have downloaded that picture and then just dropped it over here. Then let me show you one more really powerful workflow that I built a couple of months ago. So here is a secondhand shoe and I actually took this picture from my office. So this is my shoe and I connected that picture into an assistant node where I just tell the AI that I want to have a detailed product style photo of this main product photo against a gradient backdrop and yada yada yada. And then I got this picture back and then I created a couple of versions of that picture. So I named this picture main photo and then I have tagged it into the conversation.

[00:40:34]So top down view, same scene, main photo, front view, main photo, detail close-up, front part, main photo. And then I have also enhanced one of these pictures. So it is this one. I use the image upscaler node which is just a utility node that you can drop in over here. So it's just going to accept an image input and it will return a better result or better resolution and quality result as an output. And then I have also added another assistant node where I tell the AI to give me a prompt with which I can turn this picture of the shoe into a 360 rotation clip. So I didn't know how to describe it by myself. So I just asked Chachi PT directly over here and it gave me this kind of a very smooth and concise description. Then what I did was I connected this picture and this assistant into the video generator node and I hit run and I got a result like this back. And now this is super super powerful because as you can see these are all repeatable workflows. So once again you could replace this picture with an entire product catalog of for example some secondhand shoes. And you can even tell the AI to make those look like a secondhand shoes. You can actually see that in some of these pictures, it is indeed looking like it has some wear on it. So you could do this on a 100 pictures with one click of a button once you have spent like 15 minutes creating this workflow. And this is what those traditional AI image generators such as Chach or even Open Art or Hicksfield wouldn't allow you to do because you'd need to do everything manually in different windows or in worst cases even some different subscriptions. So you need to have multiple accounts or at least multiple tabs open at the same time and it would take a ton of time to run this at scale.

[00:42:22]And that is exactly why there is a platform like magnific.com. And that is exactly why I'm making this video about them as well. It's going to lay it all down inside this one canvas that is easy to follow, easy to organize, and easy to run at scale. So I really, really hope you found this tutorial insightful. And if you have any specific questions in mind, feel free to ask those in the comment section below. If you want to access any workflow that you saw in this video, feel free to check the links in the description. I have added everything over there as well. And if you have any video ideas, if you have have any workflow ideas in mind, just let me know. I'd be more than happy to cover those. But thank you so much for participating to this course and I hope you learned as much as possible. And I also hope to see you in the