This workflow effectively commoditizes high-end production by turning complex cinematography into a streamlined assembly line for digital clones. It is a masterclass in technical efficiency that prioritizes commercial scalability over genuine creative soul.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
OiiOii.ai – Viral Clone & AI Ad Skills Workflow
Added:Today, we're testing OiOi, which is a sponsor of this video. It's an AI animation and video creation platform who want to create cinematic AI videos fast. In this video, we will test a viral clone and skills features, perfect for editing your existing footage or making great product ads. Let's go in.
So, first you go on OiOi.ai. Here you have an overview of all of the tool. On the top, you have the chat box with different features like upload and image or or use a script. You can use skills in OiOi now. Like you can create story videos up to 3 minutes, character design, scene designs, and so much more.
You have the top models available like Nano Pro or on the video model Seedance 2.0 Pro. If you click on styles, you can choose from different predefined styles like a live action cyberpunk style or more 3D style. Very nice. You can also upload your own style here and use it for later production. On the assets, you can create characters and scenes perfectly for consistent video generation. On the highlights, you have video re-imagine, character design, scene design, product ad, and so much more to choose from. Let's click on the first feature here. On the right, you see some tips. You upload your reference video, select video re-imagine, replace the character, scene, or elements, review the prompt, and generate the final video. Perfect for viral cloning.
So, we have this basic video here, a dog walking in a park.
We click now on a video.
>> [music] >> We say video re-imagine, hit start re-creating. Now he's analyzing the video, analyzing the scenes, the characters, in this case just the dog, the camera movement, how the video feels to the user. And just 10 seconds later, we have a perfect analyze about our video in this endless workspace here.
Also very nice. As you can see, he analyzed a brown dog walks forward along a dirt path in a park. Dog's tail is raised and wagging slightly. We have the scene, the character description here, the duration, the cinematography, where the character, in our case the dog, is positioned. It's on the lower left quadrant. Exactly perfect description here. The visual treatment, it's a lift gamma shifted towards warm greens and earthy browns. Perfect. Sound is instrumental. Also correct. The good thing is now we can customize each step here, each character and scene. Just go on the right side here. You have the different characters, scenes, and elements. We can now replace them. Here we have a short description, but I got it. We can replace the dog. We can replace the scene, the park scene. We can replace the dirt path, the trees, the grass. It's just crazy. So now we uploaded a black dog instead of the brown dog. Replace the brown dog with the uploaded image of a black dog. Hit enter. He's now pre-generating a review matching the source files with our replaced files. And here we are. We have the prompt here for the overall video reference, only the visual style of our black dog. He's describing the scene here. If we click here, we can see he's using seed ants 2.0 fast with 8 seconds, 720p.
We can change things here like make a longer duration, or we can change the model he will use. Let's try the seed ants 2.0 pro. We can adjust things here like adding audio, video start and frames, everything to customize your viral clone here. We are fine. We just hit enter, confirm, and generate. The good thing is we can work parallel with different tasks simultaneously. So while that black dog is generating, we try the same video with a other scene. Again, we upload our dog, click on the video, reimagine, hit enter. Again, seconds later we have the shot analysis here. Go on the right, scenes, replace, add image. We have this snowy forest, replace the park scene with the uploaded snowy forest environment, hit enter. It's working so smooth in this endless workspace here, just easy. And now he's finished, and as we can see, it's a perfect viral clone here, just replacing our dog, keeping everything else consistent with the original footage here. It's just one of the best replacements I've ever seen.
Crazy good. Here we have the other example with the same dog, but a different scene. Let's look at it.
It's hyperrealistic. Just look at the feet here.
The snow is in the air. The paw is correct. He not only replaced the background, he adjusted the camera angle. The camera follows the dog through the desired scene here. Good.
Another cool feature of OiOi is skills.
You have different skills here, like a World Cup skill, a home building materials. We will try the digital display advertising. Just hit it. As you can see, the skill is automatically selected here. Create a headset display ad. We uploaded this image. Now he's analyzing our black headset here on a wooden table, gathering all information he needs. In the process, we have different options to choose from, like the model character option. It should be a active sporty woman in the ad. We can now upload our custom character here. We have this red woman. Just upload some pictures. Now we can select a background scene. We want a minimalist cozy desk.
Hit next. We can also upload a custom scene here, but we are fine with it. We just proceed. Now he's gathering all information together, as you can see. We have our sporty woman. We uploaded our custom character. We want a minimalist cozy desk environment. And only after 1 minute he's finished. He generated a full ad here. Let's look at it.
>> Meet Headzy. Designed for ultimate ergonomic comfort.
Block out the world. Immerse [music] yourself in the sound.
Headzy.
Your sound, your way.
>> This is just crazy. Oi oi just nails it here. Multi-shot scene. Our headset named Headzy. Cinematic quality. Our custom character consistent over the shots. It's just amazing. If we put this now in the viral clone, we can see we have multiple shot scenes here analyzed perfectly. Each scene with the description of it. Our goal is now to replace the black headset in this ad with a [music] white headset. So we just go in the right side. This image writing a short prompt replace the black headset with the uploaded white headset. Keep everything else the same. Hit enter.
>> Meet Headzy. Designed for ultimate ergonomic comfort.
Headzy.
Your sound, your way.
>> It's just amazing. The same environment, the same ad. A perfect clone here. Just awesome. For the next one I created a asset, a character. I upload an image here instead of prompting. It's me. And now design a character based on this image. Hit enter. He's now planning the steps here. A full automatically workflow here. As you can see the art director invites the character designer.
It's like a agent hires other agents to complete our desired task here, in our case the character designer. It's like a full team you can hire 24/7. So, now he generated a description for a surprised cat man. Perfect. We can now upload a style or choose a style fitting to the image. As we saw before, there are a ton of styles you can use. We just used the live action style like this one. Now he's generating the final character for us. And as you can see, it's just crazy good, exactly me. Perfect quality, no deformation. We can also hit multiview, so he's generating some different angles for it. And this is just crazy, the multiview angle feature here, generating our character from different angles, side view, back view, with awesome details like black baseball cap, surprised expression. All is accurate, we have a color palette, we have everything here. Also, the height is correct, that's very, very nice. On the right, we have adjustable back strap as a cap. It's still my face, just crazy good. We can now edit it or download it.
On the left you can see, "Satisfied, please continue to generate the character avatar." Or we can further modify it like we want a red shirt or something, but we are satisfied, please continue. Now we want to use this character. We have the prompt of a cinematic shot of a happy man walking along a busy sidewalk in New York City passing shops. We select our asset, our custom character here, put it in the prompt, and here is the result.
Just amazing. It's me in New York on a sidewalk, happy, cinematic quality, no deformation, consistent. Oi oi is just awesome. If you liked Oi oi, just click the link in the description to get some extra bonus credits. See you in the next one.
Related Videos
AI Agent Mastery Certification Course: Lab 4 – Tools & MCP
arizeai
350 views•2026-06-16
Real-time Voice cloning, Kimi K2.7 CODE, GLM 5.2 and 3D reconstruction | AI News
kaiexplainsYT
111 views•2026-06-16
He Believes AI Could Replace Humanity Faster Than Anyone Expects
LondonRealTV
815 views•2026-06-15
General Session by Rami Rahim-The next generation of networking: From vision to self-driving reality
HPE
108 views•2026-06-17
[PLDI 2026] Flatirons 3 - LCTES (Jun 16th)
acmsigplan
191 views•2026-06-16
Google DeepMind’s AI Halves UK Housing Planning Time
60secondsignals
467 views•2026-06-17
The Creators of Claude Code and OpenClaw don't Prompt Their Agents Anymore?!
ColeMedin
569 views•2026-06-18
Why prompt injection is AI's biggest fail
usemultiplier
1K views•2026-06-17











