A useful AI UGC product prompt specifies one product, one believable setting and one visible action. Add a 9:16 frame, camera movement and lighting, then review the result against those instructions. Our UGC generator turns three fields into an editable shooting prompt; it uses a preset writing structure, not an independent research or scriptwriting model.
Updated November 5, 2026. This guide explains our current eight-second workflow and distinguishes prompt suggestions from measured results.
“Premium,” “best” and “life-changing” are not visible actions. Replace them with something the viewer can inspect. For a bottle, a simple handling shot is more specific than “show that it is amazing.” A generated handling shot still does not prove real-world durability, insulation or leak resistance.
| Brief field | Concrete input | What it contributes |
|---|---|---|
| Product | An unbranded stainless steel water bottle | A clear subject without invented brand lettering |
| Setting | A tidy kitchen counter beside a canvas bag | Context and a useful final composition |
| Visible action | Pick up the bottle and place it beside the bag | One movement that fits a short clip |
Enter the brief in the UGC tool and choose Draft shooting prompt. Read the draft before rendering. Remove extra cuts or actions when they make the eight-second sequence too busy.
For our bottle test, we used this focused input:
A single continuous vertical handheld shot at a tidy kitchen counter. An adult hand picks up an unbranded stainless steel water bottle and places it beside a canvas bag. Soft natural daylight, realistic materials, gentle camera movement, no text, no logos, no performance claims.
See the example gallery for available outputs and the observations recorded for each run. A prompt printed in a guide is an input, not proof that the model followed every instruction.
Choose a dominant camera style: a locked close-up, a gentle handheld shot, or a slow tracking shot. Combining “static camera,” “fast orbit” and “handheld zoom” makes the intended movement unclear.
For product clips, start with the product already in frame. A long establishing shot can leave too little time for the actual action. Do not require a detailed spoken testimonial and several scene changes in the same short request.
If a detail fails, describe the failure precisely before editing your next prompt. For example, replace “bad result” with “the bottle cap changes shape when the hand touches it.” Change one major instruction at a time if you want to understand what improved.
These are untested writing templates, not additional generated examples.
A vertical close-up of an unbranded wooden phone stand on a clean desk. An adult hand places a phone onto the stand. One continuous shot, soft side light, fixed camera, no text or logos.
A vertical close-up of a plain canvas pouch on a table. An adult hand slowly opens the zipper once. Natural window light, gentle handheld framing, realistic fabric texture, no captions, no branding.
A vertical tabletop shot of an unbranded ceramic mug. An adult hand rotates it slightly so the handle faces the camera. One continuous movement, neutral background, soft light, no text.
The UGC and landscape tools share one site-wide daily allowance. Check the tool's remaining count and reset time. Drafting prompts does not consume a generation submission. Failed submissions can consume the allowance; keep the page open while a job runs and save completed results promptly.
An AI-generated creator-style clip is not a real customer's experience. Do not present a generated person as an actual reviewer. Inspect the output and confirm the applicable provider terms before choosing how to use it. For an accurate depiction of a particular physical product, a real product shoot may be a better fit than our current text-only workflow.
Open the UGC generator · See real outputs · Read the no-sign-up workflow