How to Build a Kids Animation Channel with AI Full Workflow

AI just unlocked hidden gold mine and it's probably not what you'd think. Content made for kids generates billions of dollars every year. Imagine an entire business empire built around videos like this. Until now, you'd need an entire production team. We're talking animators, voice actors, and musicians. But you can now build everything you need with AI. I'm Jack. I've spent over decade creating visual effects for many of the biggest brands on the planet. And now I'm using AI to reinvent creative workflows. In this video, I'll teach you how to create animated musical content that kids and even parents will love. So by the end, you can build your very own children's animation channel. truly believe this is one of the biggest money-making opportunities in the AI space. But the door won't stay open for long, trust me. If you do enjoy today's content, be sure to whack like on that for me and consider subscribing. All right, let's get into it. So we're going to start today's video by looking at image generation. This is where we'll create our characters and our world. Then we'll move on to animating story-driven scenes, followed by animating musical scenes, including how to generate your own nursery rhyme-style music using AI. But for anyone who's not really buying into this idea, let me show you the market we're targeting here. We can start with channel like Dave and Ava who have just shy of 16 million subscribers. I'm sure you guys would have also heard of Pinkfong. These are the team behind Baby Shark. They've got almost 85 million subscribers. And then we have Cocomelon who have 200 million subscribers. And just taking look at their videos, pretty much every single one has millions of views. These guys are making ton of money. And this is the point that I'm trying to get across. Even though you might not take this type of content seriously, it represents huge money-making opportunity. I've got two young children, so understand just how strong this content can be, especially if it's actually decent quality and teaches the child something beneficial. As you've seen, the right video in this niche can generate millions of views and make the creator lot of money. So, the first step is to build out our visuals, and to do this, we're over on Higgsfield AI. This is where we're going to prepare all of our assets that we'll animate later on in the video. The beauty here is that once we have our characters and locations set, we can actually reuse them to create any type of content for our channel that we want. And you'll notice that when start pulling up these images, actually jump between Nano Banana Pro and GPT Image 2, and that's because think Nano Banana is the best at locking in style, but then GPT Image 2 is much better at building things like character sheets, for example, and making these kind of technical edits. The first thing wanted to generate was the main character for our kids animation. So, started off by actually uploading some image references of myself. added short description and then went ahead and asked Nano Banana to remove my beard. Now, the part of the prompt that's actually doing all of the work is this really large section at the end, which is defining our style. And this is something that actually put together using Claude. So, uploaded bunch of references of kids animation that knew was really successful and that personally thought was decent quality, and then asked Claude to perform an analysis to work out why this visual style works so well, and particularly why it appeals to children. And bit further down the chat, asked Claude to pull all of this analysis together and put me together style prompt that could use. Quick tip as well, noticed that Claude was referencing existing IP, so went ahead and just asked it to remove any mention of that, and got this prompt, which is the one that actually copied and then took over to use inside of Higgsfield. And just so you guys know, all of the prompts and materials that used for this project will be available for free over on my school community. Link for that down below in the description. So, really loved this first image here. exactly the type of style that was going for. So, once was happy with this starting image, the next step was to go ahead and click on this icon here to use it as reference going forward. You want to go ahead and switch from Nano Banana Pro across to GPT image two. And to get our character sheet, we have this really long prompt here. Let me go ahead and break it all down. First, I'm describing our character once again to be really specific about how want him to look, but also what want him to wear. And then, I'm actually dropping in that same style prompt that we made with Claude earlier on. So, these first two kind of sections of the prompt are really about locking in that style and consistency. And the final section here is just framework which is telling GPT image two that want to have various different views of the same character. And the reason character sheet is so important is that it's going to show the AI video model we'll use to animate later on in the video what our characters look like from every single angle. If we don't provide this information, this is where things tend to go wrong. The AI ends up having to guess, and ultimately you waste your credits. And then, what went ahead and did is using that very first image as our reference created various different characters in the same style. Now, this prompt can actually be really, really simple. So, just said, "In the same art style with no clothes or accessories, show cute crocodile stood up. So, we also went ahead and made bear. We've got duckling, rabbit, fox cub, an owl, and hedgehog. And as you can see, ended up using the exact same character sheet process once again, so that we have these details for our additional characters, too. And something else built was this lineup here, where we have each of our characters stood next to each other. And thought this was important to define the height, especially compared to our main character here. figured it would be kind of weird if the animals were the same height as him, or even taller. It would just be not the vibe that we were going for. So, again, we're providing more context to the AI by generating this. So, here I'm referencing all of our character sheets in another pretty simple prompt to describe the lineup, and just to find that only want one of each animal to be present. And using very similar process once again, we can actually create various different locations. So, we have like nice little town here. We've got like playroom, and we've also got bathroom as well. And what's really important is that everything feels like it belongs to the same style, and ultimately the same world. And this allows us to combine our character sheets with our designed location to actually place our characters into this environment. Again, very simple prompt here to get this level of result. But, the point being that we can place any character into any environment to create just endless stories. We're building out this really diverse library of image assets that we can go on to animate to create our videos. The next step of the process is to finally start bringing our scenes to life. So, to do this, you want to head up to the top left-hand corner, and head over to use Dance 2. This is the best AI video model on the market right now and actually has been for while now and no one seems to be able to come close to what ByteDance have built. So, the first thing you want to go ahead and do is select any relevant image assets. So, for example, we want the character sheet of the boy, the crocodile, and the rabbit as well as the image we generated for our kind of playroom location. We also want to upload an audio reference of our character's voice. This means we can generate as many videos as we'd like, but the voice remains the same. To do this, pulled one of my first generations into Premiere Pro and just cut the sequence down so that it was only the dialogue left. then went ahead and actually hid the video itself and went across to export. Now, for some reason, actually kicking this out as an MP3 file just doesn't work, but if you export it as blank video with that dialogue track on there, that's what's going to function really well inside of SeaDance 2. So, we can go ahead and throw in this blank video here, which contains our voice reference. Now, the missing piece of the puzzle is our text prompt and you don't want to just vaguely describe the type of scene that you're after. SeaDance 2 needs lot of technical information to actually give you the output you want. This can be pretty complex to do. I'm going to show you method which really simplifies things. The first thing want you to go ahead and do is open up tool like ChatGPT, Gemini, or Claude in my case and we want to upload this MD file here. As mentioned earlier, this will be available on the school community. This MD file is going to give Claude really clear framework to use, meaning we can give it pretty minimal loose description of what we're after and it will give us back that really detailed technical prompt. You also want to go ahead and upload any relevant images. So, this is our character sheets as well as the location. And then we want to go ahead and describe the type of scene that we'd like. For example, show the boy teaching the crocodile and rabbit about the importance of tidying up. There's small basket in the corner. He teaches them to put their toys away when they finish playing by putting them in that basket. The video ends on high that will happy and celebrate. Mommy and Daddy will be happy. We can go ahead and send this off. And you're going to get back something like this where the scene is actually broken up into individual shots dictated by these timestamps. Claude is describing the framing and any camera movement as well as the action that's taking place and any necessary dialogue. From our very simplistic input, we've got back this incredibly detailed prompt specifically built to function with Dance 2. So, we can go ahead and copy this and take it straight over to Higgs Field and drop it in the text prompt field. quick note that my framework always adds this to the end of every prompt. It's essentially telling Dance 2 not to put any music in the generation, which just makes things easier from an editing perspective. If you're trying to stitch together multiple generations and they have music baked into them, it becomes nightmare. This allows you to add your own music. Equally, if you wanted to just let Dance 2 do its thing, you just go ahead and remove the end of the prompt like that. And what's really useful as well is we can add an additional line that says ensure the voice matches. And then if we hit the @ key on our keyboard, we can select video one, which is the blank video that contains our voice reference. This means you can reuse the reference for as many video generations as you like, and Dance 2 is going to make sure that the voice remains consistent. Let me go ahead and show you the result that got from using this exact process. Let's tidy up. When we're done playing, everything goes in the basket. Just pop it in, nice and easy. Great job, you two. Look how tidy it is now. And that keeps Mom and Daddy happy, too. So, this approach is perfect for building out scenes where music isn't really the focus. It's more about the storytelling. And let me go ahead and show you guys few more generations that put together to act as bit of proof of concept. Firstly, we have this what's in the box style video. What's in the box? Shall we take look? What do you think it could be? Kids absolutely love this type of video. You can see that the animation came out really, really clean. And this is kind of inspired by something like Miss Rachel, where it's bit of recurring segment in her content. And we can also do more educational-style content where we're actually teaching children about things like letters, numbers, and in this example, shapes. Let's look at some shapes today. Circle. Square. Triangle. Star. Heart. Those were our shapes today. See you next time. Again, really slick, professional-feeling animation. But what want you guys to focus on is that each of these generations, even though they're separate, they feel like they're part of the same world. And finally, something that my son has been learning recently, how to brush his teeth. Let's learn how to brush our teeth. Wet your brush, then little toothpaste. Brush in little circles all the way round. Don't forget the back teeth, too. Then gentle brush on your tongue. Spit it out into the sink. Rinse and you're all done. Really amazing to see the level of quality you can get by using this process, combining our image and audio references with our detailed text prompt engineered by Claude, we get result like this. But what if we want music to play much more important role in our videos? At the start of this tutorial, we were speaking about channels like Pinkfong and Cocomelon, and real driver for them is musical content. So how do we actually go about building content like this for our own channel? There are various different ways to generate music with AI, but in my opinion, the best tool for this is platform called Suno. We can actually use Suno to generate these kind of nursery rhyme style songs, combine this with our video generations to get musical storytelling content. And if you want to take things one step further, there's actually technique I'm going to show you in minute where we can have our characters lip sync along to the music. Now, you can actually use ChatGPT, Gemini, or Claude to get you set up and ready for generating your music. You can see that first asked Claude to write me style prompt with similar vibe to channels like Miss Rachel or Dave and Ava. I'm saying want it to be catchy children's music, and ended up getting prompt like this. So over in Suno, if you go over to the create tab, and then click on advance, you're going to see you've got options. went ahead and just dropped that style prompt into this field here. And next, actually asked Claude to help me construct the lyrics for our song. So I'm saying, now let's think about few different things the track could be about. It needs to be aimed at young kids, let's say between the age of 1 and 4. And did mention at this point it'd would useful to kind of cover understanding the world or children understanding their own emotions. My 4-year-old boy, for example, his emotions are just completely all over the place. He hasn't quite got to that stage yet where he can kind of keep them under control or fully understand them. So, this was particularly relevant in my case, guess. And you can see that got back whole bunch of different ideas from Claude, but its instinct was also to focus on the kind of emotional side of things. started to hone in this idea of it's okay to have little cry, and then the concept of big feelings came up. The idea here, feelings that just feel completely overwhelming, essentially. And with little bit more back and forth, once again, Claude is saying that the concept of big feelings, in terms of like emotional literacy, is the way to go. just say, "Go for it. Fire away." And we get back this brilliant section of lyrics, which just feels exactly like what you'd imagine to hear on channel like Miss Rachel or Cocomelon. So, went ahead and copied this and dropped it straight over into the lyrics section on Suno. And by hitting generate, we get back result like this. Sometimes feel happy, and want to jump around. Now, know as adults, we're kind of pre-programmed to cringe at this type of music, but when tell you that my 1-year-old was one side of the house, started playing this song, and he sprinted all the way through the house to get to my laptop and listen. I'm being completely honest. That happened just this morning. That for me was enough proof that was onto something here. So, now that we're happy with our generation, we can click on these three dots here. We can get it downloaded as an MP3 or WAV file, or we can download the individual stems. And this is what you want to do if you want to use your Suno generation to create lip sync. When you click into get stems, it's going to give you few different options here. You want to go for the middle one, which is split from mix. This is going to give you your lead vocal and then all of the instrumental tracks. And then inside of your editing software of choice, you want to go ahead and cut out the section of the song that you want to use for the lip sync. Here we just have the vocal track. We don't want the instrumental as part of this. So for example, we've got this section here. It's okay. You're not on your own. So once again, we want to go ahead and export this as blank video. Back over on Higgs Field using SeaDance 2 once again, you first want to upload any relevant image references. You can see here I've just put up the key visual for this scene. didn't feel needed character sheet cuz it's very brief generation where don't want the character to be turning around and doing lot of complex movement. also uploaded that blank video with that snippet of the vocal track. And the text prompt for this can actually be really quite simple. So we're just saying cinematic shot shows an animated character singing happily. Really important that you also add in the lyrics for the section of the track. And we can go ahead and once again tag our vocal reference that we uploaded. Now in terms of the length of your generation, you essentially want to match this to your vocal reference. Maybe just add an extra second at the end for some wiggle room. If your vocal reference is 4 seconds long for example, but then you send off 10-second generation, SeaDance 2 is going to try and fill that empty space, and that's where it'll just produce its own lyrics and music, and things just get very weird very quickly. With that being said, we've got everything set up to actually generate our lip synced video. So let me go ahead and show you the result we can get from this. It's okay. You're not on your own. As you can see, this is working really nicely, but wanted to give you guys an example of what you can actually build by stitching together multiple generations. So, we're here back once again in Premiere Pro. So, just starting off from an audio perspective, we've got snippet from the big feeling song that generated earlier on. The top track here is the vocal track. The bottom track is the instrumental, and then all of these bits of audio actually came along with the sea dance two generations that put together. So, what did was for the start of the track, we've kind of just got these scenes that are kind of relevant to the lyrics to add bit of more context and storytelling. And then at the end, we've got two examples of lip syncing, one with every single character singing along to the song, and then the example that just showed you guys where the boy version of me is getting into bed and just singing those closing lines. also added the subtitles here, which is quite typical of this type of content because it gives it that kind of sing-along vibe. So, with pretty basic level of editing, let me go ahead and show you the finished result. Sometimes feel happy, and want to jump around. Sometimes feel angry, and stomped it on the ground. And sometimes Sometimes don't know what feel at all. But that's okay. It's okay to feel this way. Big feelings come, and big feelings go. It's okay. It's okay. You're not on your own. I'm not going to lie, this song did actually end up getting stuck in my head whilst was editing this together, which is exactly what you want. And was really impressed by the overall quality of this. This is exactly the type of thing that you'd expect to find on kids animation channel. It's still just wild to me that was able to build out all of the music and visuals using nothing but AI tools. With that being said, guys, we've now reached the end of today's video. do hope you've enjoyed it. This is little different to the type of content that normally make. So, if you guys learned something today, I'd really appreciate if you whack like on the video for me and consider subscribing. genuinely think making content for this niche with AI represents massive money-making opportunity. There is definite gap in the market here, feel. I'd really encourage you guys, if you do want to try this out for yourself, to be as creative as possible. Try to avoid just putting out slop. It needs to be both entertaining and genuinely beneficial to the child because this means that parents are also going to get on board with your content. There's also lot of brain rot within this niche, so I'd really try to make something that feels like it's got thought behind it and actually has quality there. Again, thank you so much for watching, guys. Best of luck with your own projects and I'll catch you in the next video.