GPT image 2 × Gemini Omni flash
Prompt - TITLE:
Premium Cold-Pressed Fruit Juice Product Commercial Storyboard
FORMAT:
• Single-page premium storyboard
• 3:4 Portrait ratio
• Fresh beverage product campaign
• 8 cinematic storyboard scenes
• Product remains the main hero
HEADER:
• Modern premium typography
• Information cards:
- Duration: 20 Seconds
- Style: Fresh Beverage Commercial
- Product: Cold-Pressed Fruit Juice
- Audio: Refreshing Liquid ASMR
• Why This Style Works section
• Orange, green and cream aesthetic
STORYBOARD:
1. Juice bottle covered with condensation
2. Bottle cap opening in macro close-up
3. Fresh orange juice pouring from the bottle
4. Juice splash frozen in mid-air
5. Fresh fruit pieces falling around the bottle
6. Extreme macro showing natural juice texture
7. Ice cubes dropping into a glass of juice
8. Final hero bottle with refreshing splash and fruit
EVERY PANEL:
• Scene number
• Duration badge
• Camera direction
• Visual
• Action
• Product detail
CAMERA:
High-speed liquid photography, macro close-ups, slow-motion splash, condensation detail, dramatic backlight, cinematic hero shot.
STYLE:
Ultra-realistic beverage commercial, realistic liquid physics, condensation droplets, fresh fruit textures, vibrant natural lighting, premium packaging, 8K.
I made a Claude skill that animates your static ads.
One of the most overlooked ways to boost creative diversity: take your already-validated statics and turn them into video.
Built an agent powered by Claude + MakeUGC that makes it a piece of cake:
- Start with one of your Meta statics
- It analyzes it and generates two video concepts a simple animation, and a more creative one
- You approve what you like
- It animates them into eye-catching videos with Seedance
3 minutes, start to finish. The most efficient way to hit Meta Advantage+ video requirements.
Your only job is taste. Guide it, pick the best options.
Giving the whole skill away. Comment "MAKEUGC" and I'll send it to you
Claude is INSANE for ad creative strategy.
I built 5 Claude skills that handle the entire creative pipeline.
Not prompts you paste and forget. Skills you install once and they run automatically.
→ Full audience breakdown with real pain points and buying triggers
→ Ranked ad angles tied to different funnel stages
→ Production-ready copy in the buyer's voice
→ Visual direction specific enough to hand straight to design
→ Batch variations across hooks, tones, and formats
5 skills. One terminal. Finished ad creatives in minutes.
Prompts are one-time. Skills compound.
They carry context from every run. The strategy gets sharper every time you use them.
If you want the 5 skills:
Connect with me
Comment "MAP"
I'll send it to your DMs.
Repost for priority access.
I made an AI influencer who got 1,250 followers and 1M views in a week - on less than $100.
My stack:
- Image gen/editing: ChatGPT 2
- Script / prompts: Claude
- Talking videos: Seedance 2.5
- Action videos: Seedance 2.5
- Video gen: MakeUGC
Learn a skill here for free.
Cybersecurity
https://t.co/RNnGCRJYNH…
UI/UX Design
https://t.co/RNnGCRJYNH…
AI Automation
https://t.co/FRACPNwXca…
AI Video Creation
https://t.co/RNnGCRJYNH…
Video Editing with CapCut
https://t.co/RNnGCRJYNH…
Repost/bookmark and get to work!
@Dilesh2004
GPT image 2+Seedance 2.0
Prompt
10-second vertical product video, 9:16, ultra-realistic cinematic UGC style. Use the uploaded KingLucky M11 pink high-speed turbo fan image as the exact product reference. Keep the product design, pink color, digital display, buttons, and proportions unchanged.
0–2 sec: Close-up of hands holding the KingLucky M11 retail box. Smooth unboxing motion, soft natural lighting, clean pastel background.
2–4 sec: The fan is taken out and held toward the camera. The fan powers on, the purple LED ring glows beautifully, and the digital display clearly shows 150. Add subtle airflow movement.
4–6 sec: Macro close-up of the control button as the speed is adjusted. The digital display changes smoothly, emphasizing the 150-level speed adjustment feature.
6–7 sec: Side-angle shot showing the powerful turbo airflow coming from the fan. Add realistic subtle air-motion effects while keeping the product sharp.
7–9 sec: Show the fan being easily placed inside a handbag, demonstrating its compact and portable design. Smooth hand movement and realistic scale.
9–10 sec: Final hero shot of the pink KingLucky M11 standing on a clean desk beside a small plant. Purple LED ring glowing, display visible, gentle airflow, premium product-commercial lighting. End with a smooth camera push-in.
Style: photorealistic, premium product commercial, soft pastel aesthetic, cinematic lighting, realistic hands, smooth camera movements, shallow depth of field, sharp product details, 4K/8K quality, no distortion, no extra objects, no change to product branding or design. Keep all important product text and features visually consistent with the reference image.Seedance 2.0
The next step is to generate the Scene by Scene images with Flow AI using the "Agent" automation feature.
For consistent character and product, upload the Character Sheet + Product Sheet + Location & Recurring Sheet, and use all these three sheets as the reference.
Free Video Editing Resources
→ CapCut: https://t.co/JQT9IczYXG
→ Adobe Premiere Pro: https://t.co/tzfMBNgEhJ
→ Anime Video Editing: https://t.co/03YShMdVXA
Which one are you planning to learn first?
Learn a skill here for free.
Cybersecurity
https://t.co/8yvCVs1Z6l…
UI/UX Design
https://t.co/8yvCVs1Z6l…
AI Automation
https://t.co/PVKAvUdJ3w…
AI Video Creation
https://t.co/8yvCVs1Z6l…
Video Editing with CapCut
https://t.co/8yvCVs1Z6l…
Repost/bookmark and get to work!
GPT Image 2 Prompt: Frame Escape
Don’t put the product inside the ad. Make it break the ad itself
One part stays perfectly photographed. One part physically escapes the frame and the escape proves the benefit
Full prompt in the next comment ⬇️
🚨 Hermes + Claude + GPT Image 2 = machine à contenu automatisé
J’ai mis au point un workflow qui me crée du contenu IA et le poste sur TikTok tous les jours.
J’ai ciblé un type de contenu qui fonctionne très bien pour ma niche.
J’ai Claude qui me génère les prompts pour le texte de mes carrousels, ainsi que les prompts pour générer mes images.
GPT Image 2 me génère ensuite les images.
Puis elles sont postées sous forme de carrousel sur TikTok.
Tout ça automatiquement grâce à Hermes, qui chapeaute tout.
Une fois par jour, sur un compte, j’ai du contenu qui tombe et qui m’apporte des utilisateurs sans que je n’aie rien à faire.
Je suis en train de mettre en place exactement le même workflow sur un deuxième compte.
Si les performances suivent, je vais multiplier le nombre de comptes jusqu’à atteindre le moment où ce contenu ne performera plus.
BREAKING: Claude Code can now make your ads like a $10K/mo agency.
One skill. One product URL. 10 finished Meta ads in under 5 minutes.
I tested it to prove it couldn't be that good.
It was.
Distinct angles. Different funnel stages. Each one looked like a separate creative team built it.
I have hired those teams. But this was sharper.
Here is what the skill actually does once you feed it a URL:
- Pulls live competitor ads and finds the angles nobody is running
- Reverse-engineers the hooks already winning in your category
- Builds the full brief: ICP, pains, customer language, the psychology behind the buy
- Writes the concepts, the copy, and the visual direction as one package
- Ships a set built to test against each other, not to sit pretty in a deck
No dashboards. No logins. No designer queue. It runs in the terminal.
And here is the uncomfortable part for the industry:
You were never paying $4K for ideas.
You were paying for briefs, decks, revision rounds, and the Slack thread that goes quiet for three days.
I dropped this into our performance engine and the first account went from one new angle a week to five a day.
More angles. More signal. We found the winning ROAS in days, not months.
If you want the exact build:
1. Connect with me
2. Comment "CLAUDE"
I'll send it to your DMs.
STEP 3: Ask Claude to break down the script Shot by Shot, including the image prompt, video prompt, Character DNA, location, props and recurring objects.
Copy the Character DNA prompt into ChatGPT and generate the character image. Do the same for Props and Recurring objects.
Holy sh*t, this is f**king insane😳
i cancelled my higgsfield subscription for this
a free repo with 7.9K stars dropped a full AI video studio that runs on your pc
it runs on 6gb of vram, even old gpus
wan 2.2, ltx-2, hunyuan video and flux built in
no uploads, no subscriptions, no watermarks
here is how you set it up:
1. git clone the repo
2. run the one-click install script
3. launch it and generate in your browser
you will not find a FREE way to make AI Videos this year than this