UGC for TikTok Shop
The format that sells on TikTok: a creator, a product held in hand, a recommending tone. Two Reference cards, the person and the product, guarantee that neither the face nor the packaging changes between videos.
AI UGC video generator
Create UGC videos without a shoot: a consistent virtual creator holds your product and presents it to camera, voice included, in 9:16 ready for TikTok and Reels.
Interactive demo
Real workflow, real results. Pan and zoom freely.
Use this workflowThe real Influencer template workflow: multi-angle creator and product cards, a base image, then a 15-second video where she speaks. Play the video inside the node.
AI UGC ads: a virtual creator pitches your product, voice included
At a glance
- What the workflow produces
- 1 image in 9:16 · 1 video of 15s in 720p, 9:16
- Canvas structure
- 5 nodes wired by 2 connections
- Models wired in
- Seedream 5 Pro, Kling V3
- Cost of one full run
- about 459 credits, roughly €4.59
UGC (user generated content) is the most effective ad format right now, but shooting ten variants with creators quickly costs thousands. This workflow builds a stable virtual creator: her multi-angle Reference card guarantees the same face on the base image and in the video.
The magic lives in the Video node: Kling V3 generates 15 seconds where the creator SPEAKS, with the voice generated inside the video itself, gestures and lip sync included. The script is a plain Text node: change the sentence, regenerate, and you have a new ad variant.
The guide building a consistent character explains how to craft a card that holds from one generation to the next, and a video ad in 10 seconds shares script recipes that convert.
How it works
- 1
Customize both cards
The creator card (face, style, 2-3 angles) and the product card (description, photos). Swap the examples for your product: the rest of the workflow stays put.
- 2
Generate the base image
The creator holds your product facing the camera, authentic smartphone look. This is the video's first frame.
- 3
Write the script
The Text node holds what she says and how she moves. The model performs the dialogue: voice, gestures and eye contact are generated together.
- 4
Generate the 9:16 video
15 vertical seconds ready for TikTok, Reels and Shorts. Duplicate the script node to test several hooks without starting over.
Variations and use cases
The same canvas, tuned for different needs: each variation is two or three lines of prompt away.
Creatives for a Meta ad campaign
A campaign needs ten variants of the same message. Keep the person card, change the hook and the setting: each variant is a generation, not a shoot, which finally makes creative testing affordable.
Unboxing and demo
The product leaves the box, a hand turns it, the label reads. It is the most demanding shot of the format: add a macro of the packaging to the product card, otherwise the model rewrites the text on the box.
Customer testimonial
Chest-up shot, home setting, one sentence per video. Stay on the side of declared staging: presenting a generated person as a real customer is prohibited by the platforms and by consumer law.
UGC in several languages
The same creator, five markets: only the line changes. See the talking avatar when the voice has to be inside the render rather than dubbed at edit time.
Common mistakes
A product described rather than photographed
The packaging then comes out approximate, with an invented brand. The product card must hold real photos, including a macro of the label.
Fifteen seconds in a single take
The longer the render, the higher the drift risk and the more expensive it is to redo. Three five-second takes cut together cost less and are fixed one at a time.
Passing the video off as a real review
A generated creator is staging: saying so protects the campaign and the brand. Ad platforms penalise fake testimonials, not declared synthetic content.
Recommended models
Seedream 5 Pro
Kling V3Frequently asked questions
Is the voice really generated inside the video?
Yes, Kling V3 generates audio along with the image: the creator speaks with native lip sync, no TTS glued on afterwards.
Can I create my own presenter?
Yes: replace the creator card's photos with the face of your choice (generated or not), and it will stay consistent across all future videos.
Is this allowed in advertising?
Yes, as long as you follow platform rules on generated content (AI disclosure where required). Videos created on your account can be used commercially.
What does a UGC video cost?
The full canvas, fifteen seconds with Kling V3, lands around 459 credits, roughly 4.60 euros. Dropped to five seconds, the same video costs 1.60 euros.
Can I keep the same creator across a campaign?
Yes, that is what the Reference card is for: it lives in your library and is recalled with an @mention in any project. The guide keeping a character consistent covers the method.
Does the video include the voice?
No, the render is silent: the voice is added at edit time or generated in a Voice-over node. For lips that follow the text, use the talking avatar.
Do I have to disclose that the content is generated?
Yes on Meta and TikTok, which both require realistic synthetic content to be labelled in their advertising tools. The disclosure affects neither delivery nor performance.
AI UGC video generator
Use this workflowFree account, no card required. The workflow opens pre-filled in your canvas.