0:00 / 0:00

Claye Ai

INSTAGRAM · 2026-02-28Source

Prompt

GLOBAL LOCK:
Subject: A female host, mid-20s, South Asian ethnicity, warm skin tone, long wavy brown hair, wearing a cozy lavender/purple knit sweater. She sits in a home office with a professional black condenser microphone on a boom arm. Background features a dark wooden bookshelf filled with books and small plants, softly blurred.
AI Subject Consistency: A high-fashion female model, European features, sharp jawline, sleek dark hair, wearing a white luxury power suit.
Environment: High-end studio settings for AI ads; warm home office for host.
Lighting: Soft, three-point lighting for the host; dramatic, high-contrast, cinematic lighting for AI outputs.
Color Grade: Warm, saturated tones for the host; cool blues and deep blacks for the "Dior-style" AI ads.
Speech: Clear, energetic female voice, professional cadence, direct-to-camera address.

[00:00–00:05]
Visual: Rapid montage of luxury brand ads. A model with green eyes holds a perfume bottle; a red "HERA" perfume ad; a man wearing a Calvin Klein watch; a woman with flowers on her face holding perfume.
Camera: Extreme close-ups and medium shots, static with slight internal motion.
Lighting: High-fashion studio lighting, dramatic shadows.
Speech: "You don't need to hire models or designers to create brand ads anymore." (Fast-paced, hook delivery).

[00:05–00:12]
Visual: Cut to host in her office. She gestures towards the camera. A screen overlay shows the URL "lovart.ai/home".
Camera: Medium shot, static.
Lighting: Warm, soft key light from the side.
Speech: "AI can do the full photoshoot and video for you. Just go to lovart.ai and start a new project."

[00:12–00:20]
Visual: Screen recording of the Lovart.ai interface. A cursor clicks "Upload Image," then selects "Nano Banana Pro" from a dropdown menu. A prompt is typed: "Luxury studio photoshoot of a model holding my product, cinematic lighting, premium brand look."
Camera: Screen capture, focused on the UI elements.
Speech: "Upload your product image, choose Nano Banana Pro, and describe the ad you want."

[00:20–00:26]
Visual: The AI generates a photo of a model in a white suit holding a Dior bag. The host is shown in a small window, reacting. The screen shows "Edit Text" and "Model Pose" options.
Camera: Split screen: UI on top, Host on bottom.
Speech: "Lovart's design agent will generate a professional ad visual in seconds. You can adjust anything: text, background, model pose, lighting."

[00:26–00:35]
Visual: A "Before" and "After" comparison. The "Before" is warm-toned; the "After" is cool blue with dramatic shadows. The cursor then selects "Kling 3.0" for video generation.
Camera: Side-by-side comparison, then UI focus.
Speech: "And the best part? These edits don't damage or overwrite your original base image. Once your poster looks perfect, you can turn it into a cinematic video using Kling 3.0 inside Lovart."

[00:35–00:40]
Visual: The final video output shows the model in the white suit subtly moving, adjusting her hand on the bag. The host returns to full screen with a "Comment ART" graphic overlay.
Camera: Full-screen AI video, then Medium Shot of host.
Speech: "Just add a motion prompt, generate, and your animated brand ad is ready. Comment ART and I'll send you the tool link."

NEGATIVE PROMPT:
Visual: Blurry faces, extra fingers, distorted product logos, flickering lights in video, unnatural skin texture, messy background in host segments, low resolution, watermarks on AI outputs.
Speech: Robotic tone, background noise, muffled audio, lip-sync mismatch, long pauses between sentences, harsh "S" sounds.

SPEECH PACK:
[00:00-00:05]
Transcript: "You don't need to hire models or designers to create brand ads anymore."
TAKE_A: (Energetic, fast) "You don't need to hire models or designers to create brand ads anymore!"
TAKE_B: (Authoritative, measured) "You don't need to hire models... or designers... to create brand ads anymore."

[00:05-00:12]
Transcript: "AI can do the full photoshoot and video for you. Just go to lovart.ai and start a new project."
TAKE_A: (Helpful, inviting) "AI can do the full photoshoot and video for you. Just go to lovart dot a-i and start a new project."

[00:35-00:40]
Transcript: "Comment ART and I'll send you the tool link."
TAKE_A: (Direct, friendly) "Comment ART and I'll send you the tool link!"
TAKE_B: (Whispered/Secretive) "Comment ART... and I'll send you the tool link."

How claye-ai Made This Lovart Kling Fashion Ad AI Video — and How to Recreate It

This case study analyzes a high-performing tutorial by @claye.ai that demonstrates how to bypass traditional, expensive production costs for brand advertising. The video showcases a workflow using Lovart.ai and Kling 3.0 to generate "cinematic editorial portraits" and "luxury brand ads" that rival high-budget campaigns from brands like Dior or Calvin Klein. By blending a direct-to-camera host presence with slick UI walkthroughs and stunning AI-generated results, the creator taps into the "solopreneur" and "e-commerce" niches. Key visual elements include warm, professional home-office lighting for the host, contrasted with the high-fashion, high-contrast, and "expensive" aesthetic of the AI outputs. The core value proposition is clear: professional-grade marketing assets for a fraction of the cost and time.

What You’re Seeing

The video is a hybrid of UGC (User Generated Content) and SaaS Tutorial. It features a female host in a well-lit home office environment, providing a human face to the technical process. The video cuts frequently between the host and screen recordings of the AI platform, maintaining a high information density and fast rhythm.

Shot-by-Shot Breakdown

Time Range	Visual Content	Shot Language	Lighting & Color	Viewer Intent
00:00–00:05	Rapid montage of AI-generated perfume and watch ads.	Close-ups (CU), high-end editorial style.	High contrast, dramatic shadows, luxury palette.	Hook: Establish "Big Brand" quality immediately.
00:05–00:12	Host introduces Lovart.ai; screen shows the URL.	Medium Shot (MS) of host + UI overlay.	Warm, natural indoor light; clean UI.	Introduce the solution and the tool name.
00:12–00:20	Workflow: Uploading a product and selecting the "Nano Banana Pro" model.	Screen recording with cursor movement.	Bright, high-key software interface.	Tutorial: Show the ease of use (low barrier to entry).
00:20–00:26	Before/After comparison of a model holding a Dior-style bag.	Split screen / Side-by-side.	Warm (Before) vs. Cool/Cinematic (After).	Reinforce value: Show the power of AI editing.
00:26–00:35	Converting the image to video using Kling 3.0 inside the platform.	UI walkthrough of video generation settings.	Standard UI lighting.	Tutorial: Upsell the "Video" capability (higher value).
00:35–00:40	Final animated ad result and CTA to "Comment ART".	Full-screen video output + Host MS.	Cinematic motion; warm host lighting.	Conversion: Drive engagement via keyword trigger.

Why It Went Viral: The "Luxury for All" Effect

The primary driver of this video's success is the democratization of high-end production. It targets the pain point of small business owners and creators who want "Vogue-level" visuals but lack the $10k+ budget for a professional shoot. By showing a model that looks indistinguishable from a real person, it triggers a "curiosity gap"—users want to know if they can really achieve these results themselves.

From a platform perspective, the video excels in retention and engagement. The first 3 seconds use "visual proof" (the perfume ads) rather than just talking, which stops the scroll. The "Comment ART" call-to-action is a classic engagement hack; it forces the algorithm to see the video as highly relevant, pushing it to more people. The use of recognizable brand names (Dior, CK) leverages familiarity bias, making the AI's achievement feel more impressive by comparison.

5 Testable Viral Hypotheses

The "Quality Shock" Hook: If you show a 10/10 visual result in the first 2 seconds, watch time increases by 40% because users need to see "how" it was made.
The "Tool Stacking" Value: Mentioning two powerful AI models (Flux/Nano Banana + Kling) in one workflow creates a "masterclass" feel, increasing saves.
The "Before/After" Contrast: Explicitly labeling "Before" and "After" with a dramatic lighting shift triggers the brain's reward system, encouraging shares.
The "Keyword Trigger" CTA: Asking for a specific word ("ART") in the comments reduces the friction of engagement compared to "What do you think?".
The "UI Transparency" Trust: Showing the actual cursor clicking buttons on a website builds more trust than a voiceover over stock footage.

How to Recreate: From 0 to 1

Step 1: Topic Selection & Positioning

Position yourself as the "AI Production Lead." This content suits accounts focused on AI tools, e-commerce growth, or digital marketing. Your goal is to solve the "expensive content" problem.

Step 2: Character & Scene Consistency

To maintain the host's look, use a consistent filming spot. For the AI models, use specific "seed" images or detailed character prompts (e.g., "25-year-old European model, sharp jawline, neutral expression") to ensure the model doesn't change between the photo and video steps.

Step 3: Capture the "Base" Product Image

Take a simple, clean photo of a product (e.g., a perfume bottle or a bag) on a plain background. This is your "input" for the AI.

Step 4: Image Generation (The "Model" Step)

Use a tool like Lovart.ai or Midjourney. Prompt: Luxury studio photoshoot of a model holding [PRODUCT], cinematic lighting, premium brand look, 8k, highly detailed.

Step 5: Refine with AI Editing

Use "In-painting" or "Generative Fill" to change the background or the model's pose. In the video, they changed the lighting from warm to cool blue to create a "Dior" vibe.

Step 6: Video Generation (The "Motion" Step)

Import your best image into a video AI like Kling 3.0 or Luma Dream Machine. Use a motion prompt: Slow cinematic camera push-in, subtle head movement, model looking at camera, soft wind in hair.

Step 7: Edit for Retention

Cut the video every 1.5–2 seconds. Use text overlays that highlight the specific step being taken (e.g., "CHOOSE MODEL," "ADJUST LIGHTING").

Step 8: The Engagement Loop

End with a clear CTA. Use a "Lead Magnet" (e.g., a PDF of prompts or the tool link) that you send to anyone who comments a specific keyword.

Growth Playbook: Distribution & Scaling

3 Ready-to-Use Opening Hooks

"Stop paying $5,000 for brand photoshoots. Do this instead."
"I turned this iPhone photo into a Dior ad in 60 seconds."
"The secret tool big brands are using to fire their creative agencies."

4 Caption Templates

Template 1 (The Problem/Solution):
High-end ads used to require a crew of 20. 🎥 Now, it just takes one AI tool. I used Lovart.ai to create this full campaign. Comment 'ART' and I'll DM you the link! 👇

Template 2 (The "How-To" Tease):
Step-by-step: How to go from a product photo to a cinematic video ad. 🚀 1. Upload 2. Prompt 3. Animate. It's actually that simple. Which brand should I try next? Comment below!

Hashtag Strategy

Broad (Reach): #AI #DigitalMarketing #Ecommerce #ContentCreation
Mid-Tier (Targeted): #AItools #BrandDesign #Solopreneur #MarketingTips
Niche (Long-tail): #LovartAI #KlingAI #AIVideoGeneration #ProductPhotographyAI

Frequently Asked Questions

What tools make it look the most similar?

Lovart.ai for the UI/workflow and Kling 3.0 or Luma for the high-end video motion.

What are the 3 most important words in the prompt?

"Cinematic lighting," "Editorial," and "Ultra-realistic."

Why does the generated face look inconsistent?

You need to use a "Reference Image" or a specific "Character Seed" to lock the facial features across shots.

How can I avoid making it look like AI?

Keep motion subtle (avoid heavy limb movement) and use a "Film Grain" overlay in post-production.

Is it easier to go viral on Instagram or TikTok with this?

Instagram favors the "Luxury Aesthetic," while TikTok favors the "Tutorial/Hack" aspect; both work well.