Skincare video ads with AI: three Formats, one packshot
Skincare video ads can be made with AI from a packshot: a model-and-product studio film, an application or texture demo, or a before-and-after with proof.

Skincare video ads usually take one of three shapes: a studio film of a model with the product, a close demo of the texture or the application, or a before-and-after. AI can make all three from a photo of your product, but only the first two can be generated honestly from a packshot alone. A before-and-after that claims a skin result needs real evidence of that result.
With Sume, each shape is one catalog Format: a ready-made recipe you call with one HTTP request and your packshot attached. The facts below come from the Format catalog, Create a run and Runs and results docs, read on 2026-09-29. The descriptions quoted are each Format's own.
Which Format fits which skincare ad?
All three are listed as callable today at the reserved sume handle. Read one before calling it with GET /v1/formats/sume/{slug}, which returns its description.
| Ad you want | Format | Its description says it is for |
|---|---|---|
| Model and product, studio look | sume-beauty-studio | “skincare launches, makeup campaigns, beauty product reels, and clean studio brand films” |
| Texture or application demo | sume-product-usage-demo | one “real action, texture, or result”; “skincare application” and “hands-on product ads” |
| Before-and-after | sume-before-after | “beauty routines” and “product transformations backed by supplied evidence” |
How do I make a skincare video ad from a packshot?
Start a run at POST /v1/formats/sume/{slug}/runs with a key that carries formats:write. Put the brief in instruction (up to 8000 characters) and the packshot in attachments as an input_image at a public HTTPS URL; Sume fetches it when you create the run. A run takes up to 30 images, each up to 30 MB. Send an Idempotency-Key so a retry with the same key and the same body does not start a second paid run.
curl -sS -X POST "https://api.sume.com/v1/formats/sume/sume-beauty-studio/runs" \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: acme-serum-studio-v1" \
-d '{
"instruction": "Vertical 9:16 studio film: a model holds the attached serum at a vanity, soft light.",
"attachments": [
{ "type": "input_image", "image_url": "https://example.com/serum.png" }
],
"generation_spend_cap_usd": 20
}'What should a skincare video ad not show?
A generated face shows what a model imagined, not what your product did. Keep generated skin out of any claim:
- Don't use a generated before-and-after as proof of a result. Use
sume-before-afteronly with your own evidence, as its description says. - Say in words only what you can support; the video should not imply more.
- Check the label on every frame against the real package before you publish.
- For the model-holding angle in more depth, see AI model holding your product.
Can I make skincare stills for the same campaign?
Yes. The catalog also lists skincare image Formats, such as sume-serum-drip and sume-cream-squeeze, called the same way and marked not for motion; AI product photography API covers them.
How much does a skincare video ad cost?
There is no fixed price per ad. A run is metered at the rates on the API pricing page, plus a 5.5% agent fee by default, and it can never spend past its cap. generation_spend_cap_usd sets this run's ceiling, up to the platform maximum of $500; omit it to inherit the Format's cap. Every receipt reports the cap as usage.generation_spend_cap_usd_micros and the spend as usage.billable_amount_usd_micros.
When the run is terminal, artifacts[] on the receipt lists every file it generated, on failures too. These are durable media.sume.com URLs, public to anyone holding the link.
Sources
Related posts
More in Use cases
- Virtual twilight: turn a daytime exterior into a dusk photo
Virtual twilight turns a daytime house photo into a dusk shot with a sunset sky and lit windows. How to make one with AI, and what to check after.
- Safety training videos for employees, made with AI
Make safety training videos for employees with an AI presenter: one hazard per clip, your own site photos, captions, and versions in your crew's languages.
- AI album cover generator: square art at 3000×3000
Generate square album art, then upscale: Apple recommends at least 3000×3000. On Sume, generate 2400×2400 and upscale it 1.25× to reach 3000×3000.
- AI avatar for online course videos: build and update lessons
Use an AI avatar as your online course instructor: one reusable avatar, a short talking video per section, captions, and one Timeline join per lesson.
Written by Sume