AI product video ads
Turn product photos into testable video ads
For ecommerce sellers who have catalog stills but no video crew. Plan hooks and scenes, generate 6s clips, then assemble for TikTok, Reels, or Shorts.
Who it is for
DTC / marketplace sellers with product photos who need creative tests without a shoot.
What you get
3 concepts, scene prompts, MP4 clips per scene, CapCut assembly notes, clear credit costs.
What it is not
Not a talking-head studio, not auto 15–30s stitch yet, not a URL scraper.
How it works
Step 1
Upload a product photo and enter only verified benefits.
Step 2
Generate 3 ad concepts (hook, captions, 3 scene prompts).
Step 3
Render each 6s scene, download MP4s, assemble in CapCut with our notes.
vs Image to Video & Reference
- Image to Video — cheapest single-clip motion from a still.
- Reference to Video — control premium when the still already works.
- Ad Kit — multi-scene planning + clips for ad tests (this page).
Credits & limits
Model: grok-imagine/image-to-video. Duration locked to 6s per scene. I2V scene cost: 80 / 160 credits (480p / 720p). Starter $4.99 = 400 credits.
- Natural voiceover or lip-sync / talking-head avatars
- Duration other than 6 seconds, or 1080p
- Multi-clip video reference unless the active model explicitly supports it
- Guaranteed Meta / TikTok ad approval
FAQ
- What is an AI product video ad kit?
- You upload a product photo and verified facts. We generate 3 script concepts and per-scene prompts, then you render each 6-second clip with Image-to-Video. Assemble in CapCut — we do not auto-edit a full 15s master in this release.
- How is this different from plain image-to-video?
- Image-to-Video is a single clip. The Ad Kit adds brief → multi-concept scripts → ordered scenes + captions/CTA notes so you can test TikTok/Reels creatives without inventing the plan yourself.
- Do you support talking avatars or lip-sync?
- Not in this release. Captions and optional self-recorded VO only. Avatar/voice will ship only after a licensed provider is integrated.
- How much does it cost?
- Each scene uses the Image-to-Video ladder: 80 credits @480p or 160 @720p. A 3-scene kit ≈ 3× that. Failed jobs release reserved credits.
- What should I not expect?
- No URL-to-video scraper yet, no guaranteed Meta/TikTok approval, no invented discounts or reviews in scripts, no one-click 30s auto-assembly.