Free & Unlimited | How they Create Stickman Animation with Free AI

0:00Now this channel called That Guy has posted just 55 videos, but somehow they've been able to pull in around 17 million views and 230,000 subscribers. And pretty much all [music] they post is literally just stickman animations used to explain everyday concepts, and they're pulling in millions of views. Every single upload is gaining a crazy amount of traction. Now I've covered how to make stick figure animations in the past, and you guys obviously loved [music] it because the last video I made got around 300,000 views. But here's the thing, a lot of you in the comments kept telling me that [music] some of the tools I used back

0:34then are no longer free and now require a subscription. [music] And you guys wanted me to create a brand new workflow showing you how to make these animations completely from scratch using free tools. So in this course, that's exactly what I'm going to do. I'm going to show you step-by-step [music] how to create high-quality, top-notch stickman animations from scratch [music] using completely free AI tools. So anyone, anywhere can start creating [music] right now. All right, so today I'm going to show you how to create 2D stickman explainer videos using a fully [music] automated AI workflow. All right, so the first thing I'm going to do is copy my master prompt from Google

1:10Docs. As you can see, it's a little bit long, but all you need to do is copy the entire master prompt. Once I've copied it, I'm going to head over to my chatbot. Now for this process, you've got a lot of alternatives. You can use Claude AI, Deep Seek, or even ChatGPT. It's really based on your preference. But for this workflow, I'm going to be using Gemini. So I'll open Gemini and paste in my master prompt. Now this isn't just a normal prompt. It basically acts as an engine. It turns the AI into a strict production assistant that's going to guide us through the entire process step-by-step. So all I need to

1:41do is submit the prompt and wait a few seconds. All right, now the chatbot is asking us if we have a model PDF for this process. It's also asking if we want to add a voiceover to our clips or if we just want silent clips. For this one, I don't have a source, so

1:55I'll enter no source. Then I'll wait for the reply. All right, now the chatbot is asking me to choose my niche. As you can see, we've got comedy, motivational, action storytelling, educational explainer, horror, suspense, true life story, listicles, countdowns, and mixed surprises. For this particular workflow, I'm going to be working with true life stories. So, I'll just select true life story. Now, it's asking us for our aspect ratio. We've got 9 to 16, which works for YouTube shorts, TikTok videos, and Instagram Reels. So, if you're creating a video for YouTube shorts, TikTok, or Instagram Reels, you can go ahead and select 9 to 16. But, if you

2:33want to create a long-form YouTube video, you need to keep this in mind. Select 16 to 9. For this one, since we're creating a short, I'm going to select 9 to 16 and paste that in. All right, so the next thing it's asking is how long we want our video to be. As you can see, we've got 30 seconds, 60 seconds, 90 seconds, 2 to 3 minutes, and 5 minutes. For shorts, I personally think that perfect duration is somewhere between 30 and 90 seconds. Since this is just a tutorial, I'm going to go with 30 seconds. So, I'll select 30 seconds and send that. Now, just in case you want to

3:04create long-form YouTube videos, you'll want to target anything above 5 minutes. For me, around 10 minutes is that sweet spot for long-form content. You can absolutely post a 5-minute video, but if you're trying to hit that sweet spot, I'd aim for around 10 minutes. Okay, now we're moving into the actual production. We've selected our format and the AI is

3:25asking us for the idea we want to use. Now, notice something very important here. The AI strictly limits every scene to around six to seven words. And why? Because if a line of dialogue is too long, the AI video tool can cut the voice off before the clip finishes. So, the timing has to be really precise. Now, we just need to select the idea we're going to work with. But, for me, I want to explore a few more ideas first,

3:49so I'll just type more. And now it's giving me bunch of fresh ideas. As you can see, there are lots of options here. We've got things like the disappearance of Flight 19. But for this particular video, I'm going to go with the second one, the theft of Einstein's brain. So, I'll type in two and hit enter. All right, let's give it a second. And now you can see how the chatbot has divided our script. Each scene contains around seven words. And honestly, you can see how clean the chatbot is with the

4:18scripting here. For me, I'm cool with this. There's no need to make it any longer, so I'll just type no. All right, now the chatbot is asking if I already have a stickman character. For this, I don't have one, so I'll go with no. Since I'm choosing no, it's going to use the stickman character design that's already defined inside our master

4:39prompt. Okay, this is an error. I just need to retry this. What I'm trying to do here is generate our character sheet. This character sheet is going to give us a visual reference for our stickman. Now, one thing to keep in mind, we're not going to use this as the starting frame of the video. Instead, we're going to upload it as a style reference. That's going to help force the AI to keep our character

5:02design consistent across every scene. You'll see exactly how I'm going to use that in a second. All right, so now I'm going to head over to Google Flow. But before I do that, let me quickly show you some alternatives you can use for generating your images and videos. The first one is OpenArt. OpenArt is really good for creating both images and

5:20videos. Another option is Kwen AI. Kwen AI is also a really nice tool you can use. And another one is but it's also very, very good. For this particular workflow, though, I'm going to be using Google Flow. This really comes down to preference, and for me, I prefer Google Flow most of the time. So, I'll head into Google Flow. The first thing I want to do is make sure I've selected image. Since we're creating a short, I'm going to use 9:16, which is suitable for YouTube shorts,

5:52TikTok videos, and Instagram reels. Again, if you're creating long-form content for YouTube, make sure you're using 16:9. For this workflow, I'm going to use Nano Banana 2. I'll also select X2, so I get two variants and can choose the best one. And always make sure agent mode is toggled off. Now, I'll generate it and wait a few seconds. All right, after a few seconds, our character sheet is ready. The next thing I want to do is review it and choose the version I prefer. Okay, so this is our character sheet. You can see the character view, the standing view, the head-only

6:26expressions, and the different poses. For me, I'm going to go with the first one. I'm cool with this. Now, we'll head back over to Gemini. Since I'm happy with the character prompt it created, I'll enter okay. And now our character is locked. Gemini is asking if I already have my own visual beats for each scene, or if I

6:45want it to design them for me. For this, I'm going to go with yours, which means the chatbot is going to design them for us. So, I'll type yours and wait a few [music] seconds. All right, now our animation prompts are ready. As you can see, we've got all the prompts we need. Since we're aiming for roughly 3-second clips, we've got around 10 animation prompts for this video. Now that our animation prompts are ready, we're going to copy them one at a time and head back to Google Flow. Inside Google Flow, I'm going to drag in our character sheet as the reference, and then paste in the first prompt. But

7:19before generating anything, we need to change the generation mode from image to video. Since we're still creating a short, make sure the aspect ratio stays at 9:16. For this workflow, I'm going to use Omni Flash. For the resolution, the highest option we can get here is 720p, so that's what I'm going to use. Now, we originally aimed for 3 seconds per clip, but on Google Flow, the minimum duration for a clip is 4 seconds. So, we're just going to go with 4 seconds. [music] And to avoid wasting credits, I'll select one variant. Remember, each prompt is generating one individual clip. Now, all we need to do is wait. Okay, our first

7:57video is ready. Now, we need to review it. One doctor quietly stole Albert Einstein's brain. As you can see, the clip came out really well. And notice something important, there's no audio cutting. Our seven-word line fits inside the clip properly. One doctor quietly stole Albert Einstein's brain. You can hear that the entire line fits. That's exactly what we wanted. All right, so the next thing we're going to do is head back to our chatbot, copy the second

8:23prompt, and then return to Google Flow. And notice how detailed these prompts are. Also notice that each prompt is written as one large paragraph with no bullet points. That's intentional because bullet points can sometimes confuse the video generator and cause glitches. From here, we're simply going to repeat this process for the rest of the clips. We're generating the video scene by scene, so we can keep the animation clean and make

8:47sure the transitions make sense. He was only hired for the autopsy. All right, for this one, I think I might need to regenerate it. And that's something you always want to keep in mind. AI can make mistakes. You can even see that here with Google Flow. So, always double-check your generations. Yeah, I'm going to

9:04regenerate this one. Once I've got the version I want, I'll delete the one I'm not happy with. Now, I'll add our character sheet as the reference again, head back to Gemini, and copy our third prompt. There's nothing complicated about this part. We're just going to keep repeating the same workflow over and over again. Copy the prompt, head back to Google Flow, make sure the character sheet is added

9:25as the reference, and generate the clip. Keep doing that until you've created the entire video scene by scene. All right, let's give this one a few seconds. Okay, our next video is ready. Now, we'll review it and check for any glitches. Instead, he took the famous brain home. And yeah, you can see the video is coming out smoothly and exactly as

9:47planned. So now we'll add the character sheet as our reference again, head back to the chatbot, copy the fourth prompt, return to Google Flow, paste it in, and generate. Once it's ready, review it. He cut it into 200 pieces. Okay, as I always say, if you're not happy with what the AI gives you, regenerate it. Let's see if we can get

10:13something better.

10:19200 pieces. Mhm. I'm still not cool with this one. Again, AI isn't perfect, so sometimes you're going to need a few attempts, but it still makes the overall job much simpler and the workflow a lot less stressful. Now our next clip is ready, so we'll review it again. Okay, we can see the jar. So from here, just continue repeating the same steps until all your scenes are finished. This car scene is a little bit bad, so I'm going to regenerate that as well and make sure we get the background and everything else looking right. Again,

10:55the workflow stays exactly the same. Paste the prompt, make sure the character sheet is attached, generate the clip, and then review what Google Flow gives you. If there's a mistake, try again. For me, I'm going to spend a little bit of time regenerating the scenes I'm not happy with. Omni Flash actually does a really nice job here, but as I said, it still comes down to preference. Any scene you're not cool with, you can always regenerate. Okay, now we're done generating the entire video, so let's move on to the next step. Since I had to regenerate some of the clips, I renamed everything to make the sequence easier to manage. As you

11:28can see, I've got scene one, scene two, scene three, scene four, scene five, and so on. I just followed the numerical order. Now, all I need to do is put the clips together. I'll come over here and click add clip. Then, I'll add scene two, scene three, scene four, all the way up to scene 10. Once all the clips are added, it'll be much easier to download the whole thing

11:54together. Now, the next step would normally be heading over to CapCut and editing the video. But, before we do that, I'm going to quickly head back to Gemini and type next. Let's give it a few seconds. Typing next takes us to [music] step six. This is where our chatbot provides prompts for an end screen and five thumbnail variations. I'm not going to generate these right now, but they're crucial for a real upload. The end screen gives viewers something visually interesting to hold their attention until the very last second, which helps with retention. And notice something else here. Inside these prompts, the AI intentionally leaves empty space in the

12:30images. That's because we're going to add our text later during editing. We can also see our thumbnail prompts, and we've got about five different variants. Now, let's head over to CapCut without wasting any more time. I'm going to import all our clips and our background music. Then, I'll drag the clips onto the timeline one after another. Make sure you're adding them in the same order as your script. Scene one should be first, scene two should be second,

12:52scene three should be third, and so on. Just follow the flow so you don't create any confusion in the final video. One doctor quietly stole Albert Einstein's brain. Now, we just need to do a little bit of editing. When you're editing these clips, make sure everything syncs properly, and that you don't have any unnecessary gaps. Remember, we generated the video scene by scene. So, if you notice a noticeable space between two clips, and it's making the video sound awkward, you can cut that space out or slightly speed up the

13:24video to make everything flow properly. It's really just some simple editing. One doctor quietly stole Albert Einstein's brain. He was only hired for the autopsy. Instead, he took the famous brain home. He cut it into 200 pieces. The doctor hid them inside glass jars. He kept them hidden for many decades. He even drove around with the jars. Einstein's family never gave the doctor

13:52permission. He finally returned the pieces years later. The greatest mind lived in a jar. The next thing we're going to do is add our background music. Make sure you choose background music that actually matches your video's style and script. And one thing you definitely want to do is reduce the volume. You don't want the background music overpowering the character's voice. So, I'll reduce the background music volume and do a little more editing to make sure everything stays in sync. Now, I'm going to add a fade in and fade out. The reason we're doing this is simply to make the background music smoother at both the beginning and the end. We want a smooth

14:26entry and a smooth exit. That's really the reason for using the fade in and fade out here. All right, now the next thing we're going to do is generate our captions. [music] For this, I'm going to head over to text. Since our spoken language is English, you can see that English is already selected. Then, I'll click generate. Now, we just wait a few seconds for the captions to be made. All right, our captions are done. The next thing we want to do is select the caption style we're going to use. Since we're using CapCut, we've already got a lot of templates available. You can choose from

14:55different preset styles, and you can also control whether your captions appear in uppercase, lowercase, or mixed case. For me, I'm going to head over to templates. There's one particular template here that's really nice for storytelling called quick. So, just head over to templates and look for quick. Here it is. This is one of my personal

15:14favorites, so I'm going to select it. Then, I'll adjust the position of the captions. You don't want that captions blocking any important part of the video. I'm also going to reduce the size a little bit. If you want yours bigger, you can always control that using the font size. For me, I'm just going to do a little positioning and reduce the size while still making sure the captions are clearly visible. Your viewers should be able to read them properly without

15:36having to strain their eyes. All right, so that's pretty much it for the captions. Let's see how everything looks. The greatest mind lived in a jar. One doctor quietly stole Albert Einstein's brain. He was only hired for the autopsy. Instead, he took the famous brain home. He cut it into 200 pieces. The doctor hid them inside glass jars.

15:59He kept them hidden for many decades. He even drove around with the jars. Einstein's family never gave the doctor permission. He finally returned the pieces years later. So, the next step is simply exporting our video. I'll go ahead and click export. Then we need to give our video a file name. For this particular one, let's go with Einstein's stolen brain. I

16:20think that works. After that, I just need to select the folder I want to export it to. And that's it. And just like that, we've used a single AI engine to write, direct, and animate a complete, highly engaging explainer video. You can see how simple the entire workflow becomes once everything is set up. So, if you want to use this exact stick man explainer engine we used today, just comment the word prompts down below, and I'll send you the complete workflow document so you can start animating right away. And don't forget to

16:49subscribe for more AI tutorials.