Nano Banana Pro is Insane for Custom AI Motion Graphics

If you've ever tried motion graphics, you know the problem. Editors are expensive. Learning after effects takes forever, and even simple animations turn into hours on a timeline.

However, everything you just saw was made by me in a single afternoon using AI, and it cost less than my coffee.

And here's the thing, it's way easier than you think. In this video, I'll break it down step by step, the three different levels of AI motion graphics.

Starting at beginner level with a map animation that looks professionally done, then moving to an animation style used by professional editors, the kind of motion graphics that would normally take hours to create.

And finally, expert level with educational exercise animations like gym explainer videos that would take days to animate in 3D.

I'll show you the best tools for each level and exactly how I generated all of them from scratch, so make sure to stick around.

The first step in replicating this map animation is to grab a screenshot from Google Maps. You can pick any landmark you want to Big Ben in London, the Taj Mahal in India, the Pyramids in Egypt.

Honestly, whatever you like. For this showcase, we'll be using the Statue of Liberty in New York. Just search for it on Google Maps, zoom out until you see the world map the way you want and take a screenshot.

Next, we bring it into OpenArt, which is the best all in one platform for both image and video AI models.

I'll leave the link in the description if you want to try it yourself. Once you're on the site, click on Images, then hit Create Image to get started.

For this project, I'm going to I'm using Nano Banana Pro, which is currently the most reliable model for keeping your images consistent.

Drag in the screenshot we just took into the Omni reference and set the output settings to 16 by 9 and your desired resolution for the prompt.

I've already created a universal version that works for generating a clean map image for any location.

You'll find it linked in the description below. This will strip away everything that isn't part of the actual Earth, leaving just a clean, realistic satellite image ready to animate.

All you need to do is paste the prompt, upload your reference image, and click Create.

Now that we have our first image, let's move on to the second one. The process is similar, but this time we want a closer zoom on the location.

Take another screenshot from Google Maps just like before, bring it back into nanobanana Pro and use this new universal prompt.

This tells the AI to do the same cleanup as before, but now it also highlights Your chosen location in red and adds a single clean label.

Everything else stays realistic and untouched. And just like that, we have our second image. It's almost perfect, but it's missing a few details and that cinematic wow factor.

What we can do next is to bring this image back in Nanobanana and put it into the Omni reference. Then we improve it.

Using this prompt. Apply a low angle tilted earth perspective to the map, creating depth and curvature while preserving geography.

Place a realistic 3D Statue of Liberty model at the correct New York City location on the east coast of the red United States region, slightly protruding above the surface.

Maintain clean satellite textures, cinematic lighting, soft shadows, no extra label.

And that's it for the image preparations. Now we get to the fun part, bringing it to life with animations.

First, for this let's head to the video section. Click on Image to Video and select Halo O2. You can experiment with other models like cling 2.6, Google VO3 or Sora 2, but I'm using Halo for its first frame and last frame feature which gives much better control over the motion.

First we add in the first frame which is our initial generated image and then set the last frame as the second generated image after that for the video duration and decide whether you want six or ten seconds.

Once everything is ready, our prompt is going to be a zoom into the United States on the map as the country becomes outlined and turns red in a stylized way with the text United States popping up over the land.

Then a realistic 3D Statue of Liberty pops up dynamically from the New York City location on the East Coast.

Hit Generate and that's how you get this exact result. The best part, you can swap out the country and landmark in the prompt to make it entirely your own.

Alright, now for the advanced level. This is where AI can actually save you hours if you're an editor.

Editor style. Motion graphics like this aren't just flashy, they're practical. Let me show you exactly how I did it.

Head back into OpenArt and go to Images. Pick your model. I'm still using nanobananapro since it's been the most reliable for me.

Keep the same output settings as before with your preferred resolution. For the prompt I use this dark cinematic gradient background you used in Video Editor Motion Graphics.

Three soft rounded rectangular text boxes with subtle highlights. Top left reads Nano Banana Pro. Top right reads any AI video generators.

Bottom center reads Custom Motion Graphics. Thin white curved connector lines extend from the top boxes and merge into the bottom box.

Centered, balanced layout, clean UI design, no icons, no Extra text. This gives us the base image. The best part here is how flexible this prompt is.

You can change the text, adjust the gradient, or move the boxes around to get exactly the layout you want.

Now we need a starting frame. Bring this image into the Omni reference. Then use this prompt. Remove all foreground elements, including text boxes, connector lines, light streaks, glow effects and UI shapes.

Preserve only the original dark gradient background exactly as is. Do not add or change anything else.

The goal here is simple. We're isolating the background so we can use it as the start frame for the animation.

This is perfect. Now let's animate it. Go to video and click on Image to Video and select any model you like.

I'm using Halo again because it supports first frame and last frame, which gives much better control.

Set the first frame as the clean background image and the last frame as the image with all the text boxes and connectors.

For the animation prompt, this is what I used. Start on a clean gradient background. The top left and top right text boxes fade in and slide gently into place.

First, thin curved lines. Then draw downward smoothly from both top boxes, converging at the center the bottom box.

Custom motion graphics fades in last and settles softly smooth, easing Professional editor style motion.

Hit generate and this is the result.

The cool thing about this is that you can completely change the animation just by tweaking a few words in the prompt.

For example, this is another version.

Just by changing a few words, you get a completely different animation and this is still just advanced.

Let's step up to expert level with educational exercise animations. So first think about the character you want to use for these educational animations.

Since this is exercise content, the goal is to clearly show muscle structure without distracting details.

I already have a character prepared, so I'll show you how I generated that. To save some time inside OpenArt, go to Nano Banana Pro.

I'm going with a muscular mannequin as the base character. The idea here is to generate a clean, neutral body that shows muscle volume, but without skin texture, facial features or unnecessary detail.

This keeps the focus on the muscles instead of realism. Nanobananapro generated this mannequin body with athletic proportions and a smooth single color surface which makes the muscle groups easy to read.

The neutral upright pose works well as a base for most exercises, and the studio lighting with a dark gradient background helps everything stand out clearly.

This gives us the perfect hero shot for educational animations where the focus stays on the movement and the muscles.

Next, bring this hero shot into the Omni reference and slightly refine it into a close up exercise setup.

The goal here is to frame the arm, simplify the scene and introduce a subtle internal muscle glow so viewers immediately understand what's about to be trained.

The result is a clean, focused image that's perfect for animation. Now we bring this into cling 2.6 to bring it to life.

I'm using cling because it gives smooth, controlled motion that works really well for fitness visuals.

For the animation, the idea is a slow, controlled dumbbell curl. The arm moves smoothly from rest into the lift while the internal muscle glow stays visible and gradually intensifies as the motion progresses.

And just like that, you've got a professional looking educational exercise animation fully generated with AI.

Here's another example. This one was created using the exact same way I reused the same hero character in Nano Banana Pro.

But change the angle and the exercise setup. You'll see the prompt on screen, but the key idea is simple.

We're switching to a rear angle and isolating a different muscle group, focusing on the upper back. There's no lifting yet, no contraction pose, just a neutral starting position with a subtle internal color change to show which muscles are about to be activated.

With this generated image, we bring it into CLING and animate it again. You'll see the motion prompt on screen, but the goal here is controlled realistic movement.

The bar moves smoothly from shoulder level to overhead, while only the targeted upper back muscles change color to indicate engagement.

So there you have it. Everything from beginner to expert for creating motion graphics using AI. All the tools and prompts I used are linked in the description below, so you can check them out and start making your own.

If you want to learn how to make viral AI shorts, click the video on the screen right now. I'll see you in the next one.