CADABRA AI

UGC product review

A real-looking person holds the product and talks to camera, as if on a phone.

What you give
A product photo, and a presenter photo if you want one.
What you get
A 15 to 30 second video: hook, one claim, proof, and the product in hand at the end. It looks filmed on a phone, not produced.
Price
Planning is free. Every paid step is priced before it runs.

The brief you paste in the chat

Make me a UGC review video of my product that looks filmed on a phone, not produced. The product: [what it is] (a clear product photo is attached) The presenter: [an attached photo of someone who agreed / or invent a new person] The one claim: [what it did for them, in a customer's words] Length and format: [15 / 30 seconds], 9:16 Language: [Hebrew / English] Second one: the product close to the lens and one customer line, no "hi everyone". The end: the product in hand, readable, and a call to action. Show me the script first. Show me the plan and one itemized price first, and start only after I approve.

What Cadabra sends to the engine

seedance-25: Keep the exact product from the reference images. Same colorway, logo placement, silhouette and materials. Do not invent a sibling SKU. Do not beautify the print into a different design. Keep the same real person as the face reference. Same age, bone structure, hair, skin. Not a lookalike. Not a celebrity. The face does not move: hands, body and camera move around it, no new expression. The presenter lifts the product toward the lens in one motion, product fills the lower frame, then settles. Shot on a phone held at arm's length, 26mm equivalent, natural indoor daylight from a window, slight handheld sway, real room, not a studio, true skin texture, no beautification.

How it works

Use it when you want an ad that looks like a customer talking about the product, not a production. That is the whole UGC form: footage that looks filmed by a person on their own phone. The first second is the product close to the lens and a customer line, never a greeting.

How it works: the presenter comes first. If you attach a photo of someone who agreed, it is used, and with more than one shot the face is locked as a character so it stays the same face. If not, a new presenter who resembles no celebrity is made, and you see the picture before anything is animated. Then a script is written: one claim, in the words of a customer, followed by proof. The video is built from three shots: a hook where the product is lifted to the lens, the claim to camera, and a close-up of hands using the product.

At the end it is cut into one video, with captions on every spoken word, because most viewing happens with the sound off, and an end card with the product name and one call to action. A tip: write the claim the way a customer would say it, short and plain. A claim that sounds like an ad breaks the whole spell.

Questions and answers

Can I use a photo of anyone?

Only of someone who agreed. Without one, Cadabra makes a new presenter who resembles no celebrity and shows you the picture before anything is animated.

How does Hebrew speech sound natural?

In Hebrew, the voice never comes from the video engine. The lines are written at about three words a second, recorded as a voice, and the talking shots are lip-synced to that recording.

What does it cost?

Planning and the script are free. Every paid step, the shots, the voice and the edit, is priced before it runs.


עמוד הבית · בלוג · וידאו AI · איך עושים סרטון AI · מוזיקה AI · תמונות AI · קול AI · תמונה לסרטון · כתוביות בעברית · כמה זה עולה · אפקטים לעריכה · פריסטים · כל המנועים · קרדיטים ולא מנוי · AI בעברית · קורס מתחילים · מדריכים · נגישות

Cadabra AI · סלע מייטל (עוסק פטור) · Meital Sela · רחובות