15 AI Video Styles for Social Media Using Google Flow | Karan Rathore | Skillshare

Playback Speed


1.0x


  • 0.5x
  • 0.75x
  • 1x (Normal)
  • 1.25x
  • 1.5x
  • 1.75x
  • 2x

15 AI Video Styles for Social Media Using Google Flow

teacher avatar Karan Rathore, Creative Technologist: Video Editing, AI, and 3D

Watch this class and thousands more

Get unlimited access to every class
Taught by industry leaders & working professionals
Topics include illustration, design, photography, and more

Watch this class and thousands more

Get unlimited access to every class
Taught by industry leaders & working professionals
Topics include illustration, design, photography, and more

Lessons in This Class

    • 1.

      Introduction

      1:48

    • 2.

      Interface Tour Google Flow

      1:19

    • 3.

      Understand the Video generation prompt

      2:23

    • 4.

      Style Spotlight 1990s Hand Drawn Anime

      2:04

    • 5.

      The Face Swap Trick

      0:48

    • 6.

      Timed Scenes and Sound Control

      0:53

    • 7.

      Live Action + Cartoon Cutaway (Flat 2D)

      0:50

    • 8.

      Vinyl Toy Figure

      0:46

    • 9.

      Low Poly Horror

      1:28

    • 10.

      16 Bit Pixel Art

      1:06

    • 11.

      Claymation

      0:48

    • 12.

      Paper Cutout Pop Up Book

      0:58

    • 13.

      Scribble Action Comic

      0:53

    • 14.

      Oil Painting

      0:41

    • 15.

      Synthwave Retro

      0:50

    • 16.

      Realistic Fantasy Transformation

      0:35

    • 17.

      Collage Art

      0:20

    • 18.

      Disney Pixar 3D

      0:34

    • 19.

      Y2K Tech Retro

      1:04

    • 20.

      PS1 Retro Game

      0:37

    • 21.

      Project

      1:01

  • --
  • Beginner level
  • Intermediate level
  • Advanced level
  • All levels

Community Generated

The level is determined by a majority opinion of students who have reviewed this class. The teacher's recommendation is shown until at least 5 student responses are collected.

25

Students

1

Project

About This Class

Stop scrolling past AI videos that all look the same. In this class, you'll learn 15 distinct, scroll-stopping visual styles you can generate in Google Flow using Gemini Omni Flash - each one broken down prompt by prompt, so you're not just copying a template, you actually understand why each one works.

What you'll learn:

  • How to read and build a Google Flow prompt using the 3-layer framework: Identity, Style, and Action — the difference between a prompt that gives you the same great result every time and one that's a total gamble
  • 16 ready-to-use visual styles, including retro 90s cel animation, claymation stop-motion, synthwave neon, vintage travel-poster collage, low-poly horror, and live-action-meets-flat-2D composites
  • How to keep a character's face and identity consistent across multiple clips, so you can build a recurring character for a content series instead of a one-off video
  • How to write prompts that reliably pass Google's content policies, so your videos actually generate instead of getting rejected
  • Real hook and pacing techniques for turning any of these styles into content that's actually built to be watched, not just pretty

Who this class is for:

Anyone making content for TikTok, Reels, or Shorts who wants their AI videos to look intentional instead of generic - total beginners who've never opened an AI video tool, and intermediate creators coming from tools like Midjourney or Adobe Firefly who want to add motion and consistency to their toolkit. No coding, no editing software required to get started.

Materials/Resources:

A Google Flow account (the free tier is enough to follow along), one or two reference photos if you want to build a consistent Character, and the full downloadable prompt library and style cheat sheet included with this class.

By the end, you'll have a personal library of 16 tested prompt structures you can adapt to any niche — and, more importantly, you'll understand the pattern underneath all of them well enough to build your own style 17 yourself. Let's get into it.

Meet Your Teacher

Teacher Profile Image

Karan Rathore

Creative Technologist: Video Editing, AI, and 3D

Teacher
Level: All Levels

Class Ratings

Expectations Met?
    Exceeded!
  • 0%
  • Yes
  • 0%
  • Somewhat
  • 0%
  • Not really
  • 0%

Why Join Skillshare?

Take award-winning Skillshare Original Classes

Each class has short lessons, hands-on projects

Your membership supports Skillshare teachers

Learn From Anywhere

Take classes on the go with the Skillshare app. Stream or download to watch on the plane, the subway, or wherever you learn best.

Transcripts

1. Introduction: We see them all over social media every single day, those mind blowing high retention AI videos that make you instantly stop scrolling. The second of unique visual styles that look incredibly complex, leaving you amused, inspired, completely clueless as to how they were actually made. I'm current a creative specialist and online instructor. If you ever tried jumping into the world of AI Video generation, you already know the struggle. You open up a tool, look at the blank prompt box, and get completely overwhelmed. By all the settings and styles, you end up confused, wondering exactly what to type to get those premium viral quality results. The camera punches, that's exactly why I build this class. Instead of wasting hours guessing, we are going to master a single incredible powerful tool, Google Omni. In this class, we are going to break down the exact trend setting styles dominating the Internet right now. We will explore Retro 90s cartoon, claymation, pixel art, commercial, jurism and low poly horror ethics. I'm going to show you exactly how to structure your prompt to generate them simussly. By the end of this course, you won't just copy prompt. You will understand the core mechanics of the visuals to build your very own signature video style for your brand or client project. The tool are ready, the workflow is streamlined and your creativity is the only limit. If you are a freelancer, graphic designer, a video editor, just want to know how to generate a content for your social media page. You can do it with Google Omi and I will show you how you can do it. Let's start generating a first insane AI video with Google Omni. Let's get started. 2. Interface Tour Google Flow: Before we write any prompt, let's look around first. Think of this like a tour of a new house. Go to labsgogle.com slash FLO OJ Search Googleflo. It will take you there. Sign in with Google account. That's all you need to do. At the top of the skin, you will see a few things, your project name, the video shape, and which AI model you are using. You will also see your credit balance. That's how many videos you can still make. Don't have a paid plan. Don't worry, you still get free carts every day. That's enough to make two or three videos a day. Great for practice. Now click New Project to Start. Inside, you will see a space for ingredients. That means the reference images and characters you want to reuse. In the middle of the screen, is the big prompt box. This is the heart of everything. Every idea we teach on this course gets tied right there. Once you click Generate, your video appears in a grid. You can drag any video into a scene Builder. If you want to connect many clips together, that's a whole workspace in one quick tour. Now that you know where everything lives. Now let's learn how to write a winning. 3. Understand the Video generation prompt: Here is the biggest secret in this whole course. Every strong prompt has three layers. Most beginner only use one layer and got confused result. Let's break it down using a real example. Layer one is the identity lock. Example, keep the same character sheet and the same framing from this image. This does two things. Most people only notice one. First, it locks the face and body to your reference photo. The AI won't guess a random person. It uses your photo. Second, it locks the camera angle too. If you skip this part, you may still get the right person, but the camera shot might change completely. Always say keep this exact shot. Not just keep this person. Layer two is the style layer. Example, restyle this a 90s hand on two D animation. Use thick black ink outlines, flat cell shading, slide line wobble and light film grain. Notice how specific that is. It's not just make it animate. It's too vague. Each phrase tell the AI one exactly drawing trick. Ink Outlines means bold lines around every character. Line wobble. Copies, how are human hands slightly shake when drawing by hand. Film grain at that old scratchy VHS tape. Filming Bing this specific is what separates amateur prompt from pro prompt. Layer three is the action layer. Example, the character is moving through a subway and fighting. This tell the AI what is actually happening in the scene. Notice the character get mentioned again here too. That's not wasteful. It's remained the AI to keep the same identity. This action layer is a little open ended on purpose. It doesn't say the exact camera angle or fight move. Sometimes that's good. You let the AI be creative. You know this. The action layer is where results change the most between tries. Your identity and style will usually stay steady. The action is the person part is still open to surprises. So remember these three layers, identity lock style layer action layer. Master this formula, and you can build almost any style in the course. 4. Style Spotlight 1990s Hand Drawn Anime: Use it, drag your reference photo into the engaging box. Click at two prompt. You photo is now linked let's put the formula to work with our first style. We already saw pieces of this prompt in the last lecture. The full look is 90s andra on to DSL animation. Key details thick black in outline, flat color shading, slight color bending, a little line vowel and light film grain over the whole video. Paste in the style prom, describe the action, then it generate. Wait about one or 2 minutes and your retronim video is ready. After pasting your prompt into the prompt box, you can add the characteristic or the exact reference where you want to show your character. Like this, you simply have to press and all the option that you have uploaded will show in the prompt, and you simply have to click that and add to prompt. That's all you have to do. After that, if you want to enable the agent, you can enable the agent. In the video section once you click here, you can use either one to generate the image or either one to generate the video. Right now we are selecting the video and the ingredients. If you want to frame, you can check the framing and the shape of the video. If you want the real style 1916 ratio, you can use that or if you want to use it for the normal video, you can use that. After that, you have to select one oh two generation. It depends on your users. If you want to generate two video at one time, you can do so. And about the model, which model do you want to use, you can use that. And the, the duration of the video, you can select that also. After that, you simply have to click Generate. The generation will use 15 carats. So this time it takes only 15 cares for ten second generation. 5. The Face Swap Trick : This prompt, I have used my face again, once you click here, and you will see here is the face. And this is the prompt that I have used. And this is the output. This is my scene. You stole it from me. So this is the output that we got, right now. So this prompt uses a hybrid composite technique. It forces the AI video engine. Again, the blueprint, it's at the video format, the separate the words that is scene versus character, the style that we wanted. Here is the video that has been rendered with my face. It has added new characteristics, and now I am in the cindrella. It looks nice to 6. Timed Scenes and Sound Control: Ready for advanced version of prompting. Now we add two powerful new controls, Control number one, time sequence. Instead of letting the AI animate randomly for ten second, you set exact timestamp example 0-6 second and keep it realistic. From six second to ten second transform into a cartoon character. This keep the switch smooth instead of a messy sudden change. Control two native sound direction. You can tell the AI what sound to add right inside your prompt. It gives you more control over the finished video fields. Together, these two tools help you build short stories, not just single images in motion. What? 7. Live Action + Cartoon Cutaway (Flat 2D): Live axon part, built trust and pacing, while the style you want is a crisp hard cut between the two visuals rather than a smooth fit. Here is another video that has been generated. Mm hmm. It is a sticker composite. Imagine a complete photorealistic real world, but a flat, thick outlined cartoon sticker character is placed directly inside it. This style keep the two visuals medium completely separated in the same frame. The real world scene handles the light, shadow, and team naturally, while the character animation stays 100% flat with zero gradient or shady, creating an eye catching contrast. 8. Vinyl Toy Figure: Now let's step into the T D world with the vinyl til aesthetic. This look mimics high and designer collectival figure. The skin textures are completely smooth with no visual pores showing a glossy plastic or vinyl sheen. Instead of individual stent hair and clothing are rendered in heavy sculpture gloms. The light behaves exactly like a professional studio and a product photo, catching the rounded edge with sharp, intense bright spot. So here is the video that I have created. Take a look at it. 9. Low Poly Horror: Step I follow to create this video is the same. I use this as a reference and use the prompt. You will find the prompt in the material section, so don't worry about the prompt. So now let's move to the next that is low poly three D art. This is the saved frame, and this is the video that has been created using the frame. It looks scary at the end. So it features a minimal geometric look polarized by early Indie walking games. And character and objects are built out of basic polygon shape with sharp edges and no smooth color grading. The surface are entirely met and non reflective, lit by flat, dim ambient light. This blocky simplified look feels a little unnatural, which makes it an incredible style for creating atmospheric unease or horror without needing cure. Here is another video that I have created, and it looks really amazing. And this is a prom that I have used. And you will find it in the materials section. Okay, leave. It looks scary, just like the PS one games. 10. 16 Bit Pixel Art: We go back to one more generation, we find a 16 big pixel art cuts in style. This look relies on a visible pixel grid with absolutely no smooth shading loops, high silte value and deep color contrast, such as a warm fire lightning glowing against a cool per night sky depth is built using layered paraac scrolling, where the foreground elements slide by faster than the distance background structure. Take a look at it. At the end, it added a helmet, and this one is different from this one. I don't know how this get here, but every time you create video from Omni, it just give you at the end an error or a glitch or it just thing on its own. Till 8 seconds, it's working fine. 11. Claymation: Look at deeply technical style claymation stop motion. In this aesthetic, everything looks like a hand crafted real clay, painted wood, dried moss and crumpled paper. The defining detail here is the organic imperfection. You can actually see the tiny fingers and tools marks on the surface. The lightning behaves like warm lamp on a small miniature studio, creating a shadow depth of field that keeps the focus sharp on our main subject. Let's watch this video and you will understand what I'm talking about. Hmm. 12. Paper Cutout Pop Up Book: Next, the vinyl paper cutefact which looks like an animated mid central travel book from 1990. It's combined realistic magazine photo cutout with hand drawn paper shapes. The visual properties feature rough torn paper edges, vintage printing, half ton dots, and plenty of clean or white empty space. The animation style mimics a physical pop up book where flat paper layers flip, fold, and snap into place smoothly, and this is the video that has been created. It looks like someone has created this video from its hand, but it is AI generated. 13. Scribble Action Comic: Now for intense kinetic action, we use rose crabs crash style. This is a loose messy hand drawn look that uses frantic black and white ink lines and quick cross hatching for shadows, the strict rule of the aesthetic is that it remains completely black and white using keb ink line for energy and smoke. The only exception is a single flat neon color, splash is reserved for a major final impact frame to maximize visual shock. Here's another output that I have tried. 14. Oil Painting: Looks very much alive. Now, if you want to relax artistic feel, the impression oil painting style, turn the entire video into a moving canvas. It replicates thick, visible brushwork or soft blended oil painting structure with absolutely no hard digital lines. Light diffuses softly through the oil texture, and the scene shifts seamlessly at the camera glides toward the single continuous tick watching this video, and you will get what I'm talking about. Oh 15. Synthwave Retro: Electric throwback, the synth wave retro Illustration Style, inspired by a 90s 80 retro futurism. It uses bold cell shaped outline over deep purple and Mcenta light, sky framed by a zit horizontal sun and neon city SiludeT anchor the classic media feel. Every single frame is coated in a heavy analog television half tone grain texture. Let's play with this video, and you will see the Synth wave. Sometimes the helmet come and sometimes the helmet goes. This is the glitch. You always get all the Google video model. 16. Realistic Fantasy Transformation: Let's take a look at the realistic transformation style, which blends normal real world environment with pure fantastic logic. The realistic rule of this style is the unbroken tracking shot. The camera moves forward continuously with no editing cards, allowing one location to originally dissolve and shift into a complete, different massive setting right before your eyes. What a sensational shot. 17. Collage Art: Is the style I sometime love to see. As you can see, all these are the cut out from the kibble of the day and has been added into the ide. 18. Disney Pixar 3D: Other style are Disney Pixar style. This is a cinematic theory style where it replicates big budget modern theatrical the animated feature film. Let's play this one. Focusing entirely on the subject. Some parts look really awkward that it has generated randomly, but I wanted to explain and share with you it conveys really good. 19. Y2K Tech Retro: So this is a white two K to futuristic style where I have used different images using all the prompts and my face as a reference and created using nano banana. And after the late 90s early 2000 tech bubble culture and media packaging design heavily utilizes resendent chrome coating, translucent, neon plastic and bright silver metallic finishes, cent with vintage digital gadgets like pagers or cybernetic accessories. The overall visual is treated with extreme contrast bright white bloom around light source and sub analog video signal flows to give it a nostalgic early digital crit by combining all the frame I have generated. You have seen this video or this type of AI trend on your screen before. 20. PS1 Retro Game: Deep gaming nostasia, we have the PS one era survival horrors type of late 1990s retro video game. This is defined by low resolution screen, muddy textures, visible pixel artifacts, and affine texture wrapping, which cause texter to zitter when the perspective move to finish the vintage. Look, the entire frame is given a subtle CTR television screen line freak. 21. Project: Congratulations. You have officially made it to the final lecture of this course. We have covered a lot of ground from learning how to separate our background and character using strict prom block, different completely unique animation and design styles. Now it's your time to take everything you have learned off the screen and put it into practice. For your final class project, you are going to pick just one style from our reference guide and create your own AI video using Gemini Omni and Google Flow. Your very first step is to open up the course resources folder and look through our style cheese Sheet. Pick the exact look that excites you the most. Pick your style first because that choice dictates the visual language you will use. I cannot wait to see the incredible style and variation you all came up with. Thank you so much for joining me in this class. Experiment push the boundaries of your prom, keep creating and I will see you in the next course.