Subscriber Perk

Get 50% Off MiniMax H3 and Other Selected Video Models

P—12 / UGC OVERSIZED PASTRY

UGC video ad prompt with lip sync for MiniMax H3

Five shots, five spoken lines, one absurdly large croissant.

A creator-style bakery video: shot on a phone by a friend, five shots, and a line of dialogue in every one of them. It is the hardest lip-sync job in the library — five separate deliveries across fifteen seconds — and the prompt manages it with one rule applied five times: the mouth has to be visible when the line is spoken.

Steps: 2 steps
Mode: Image to Video
Duration: 15s
Ratio: 9:16
Resolution: 2K
Audio: Dialogue + street ambience
Cost: $1.95

Reference render — not generated on this site. Source: pollo.ai

Output reference

A creator in a wide black hat standing on a city pavement — Step 01 — the still

A creator holds an oversized croissant outside a bakery and speaks to camera — Step 02 — the clip

JPG · PNG · WebP · HEIC · 30 MB

Create a UGC video ad featuring [image1], presenting product [image2].

15-second vertical 9:16 authentic UGC bakery sharing video, shot on a phone by a friend, bright natural daylight, casual handheld micro-shake, photorealistic, real skin texture, exact English lip-sync, no subtitles, no text overlays, no watermark.

Use the reference person: stylish young Black woman with dark skin, braided hair styled into two low buns with long face-framing braids, wearing a wide black hat, black sleeveless fitted top, white mini pleated skirt, white collar detail, black-and-white belt, soft confident expression. No handbag.

Background: a cosy modern bakery storefront, warm beige facade, a big glass window full of pastries, outdoor wooden tables and chairs, soft golden afternoon light, and an oversized croissant sculpture above the shop.

Hero food: an oversized giant croissant, almost half her body size, golden brown glossy crust, flaky layered ridges, buttery texture, crisp edges, warm bakery shine.

Shot 1: medium full-body outside the bakery. She holds the giant croissant with both arms, laughs at how huge it is, faces camera with mouth visible and says: "Okay, this is officially the biggest croissant I've ever held."
Shot 2: close-up of the croissant. She tilts it toward the camera to show the glossy crust, flaky layers and huge size, then says: "Look at these layers. It's literally bigger than my bag."
Shot 3: she takes one real bite from the edge, crunch sound, crumbs fall naturally. She lowers the croissant so her mouth is visible, reacts with widened eyes and says: "Wait. It's actually so buttery and crispy."
Shot 4: she hugs the giant croissant playfully, standing in front of the bakery window, and says: "I thought this was just for photos, but no, I'm eating this."
Final shot: full-body bakery exterior, giant croissant visible in her arms, soft smile to camera: "If you see this place, get the giant one."

Natural street ambience, bakery cafe sounds, crispy bite, light crumbs, live on-camera speech, exact lip-sync. The giant croissant stays flaky, golden and clearly visible throughout.

2121 / 7000

Images 0/9 · Clips 0/3 · Audio 0/3 · 12 files max

Sign in free · 3 clips a day

No card, no account: watch and download 12 real MiniMax H3 clips, and run one prompt of your own on Agnes Video V2.0 — 16:9, watermarked, once. After that, sign in free for 3 clips a day on the free engine. MiniMax H3 on your own prompt starts at $24.9 a month.

1 FREE CLIP · WATERMARKED1 FREE CLIP · Agnes Video V2.0 · 16:9 · 5s · WATERMARKED

02 — Inside

Two prompts, one workflow

Each step is the result beside the text that produced it. Copy either prompt into the console above.

Reference output, step 01

The still

A creator in a wide black hat standing on a city pavement — The still

Step 01 — the stillAny image model

1,417 chars

The creator

The face and wardrobe that has to survive five shots. Everything gestural is deliberately left out so the video pass can pose her.

Photoreal candid street-style portrait, vertical 2:3, natural daylight.

SUBJECT:
A stylish young Black woman in her twenties with deep skin tone, standing straight on to camera with a calm, quietly confident expression. Her hair is braided and gathered into two low buns with several long face-framing braids hanging loose past her collarbone. Soft natural makeup, no heavy contour.

WARDROBE:
A wide flat-brimmed black straw hat with a ribbed crown. A fitted black sleeveless knit top with a small scalloped hem, cropped at the waist. A short white pleated skirt with a double black eyelet belt at the hip. A narrow white collar band at the throat with a thin black ribbon tie. No handbag, no jewellery beyond the collar.

FRAMING AND POSE:
Full body, head to mid-calf, subject centred and occupying the middle of the frame. Camera at chest height, straight on. Arms relaxed at her sides — no gesture, nothing held, hands empty and visible.

SET AND LIGHT:
A quiet city pavement beside a pale board-formed concrete wall, low green planting behind her, a blurred modern building further back. Soft overcast daylight, gentle contrast, no harsh shadow on the face.

RENDER:
Photoreal, real skin texture with visible pores, natural fabric weave, sharp on the face, mild background falloff.

RESTRICTIONS:
No text, no logo, no brand marks, no watermark, no second person, no beauty-filter smoothing, no exaggerated pose.

Reference output, step 02

The clip

A creator holds an oversized croissant outside a bakery and speaks to camera — The clip

Step 02 — the clipMiniMax H3 · Image to video

2,121 chars

The UGC take

Five shots, each one naming a camera distance, an action, whether the mouth is visible, and the exact line.

Create a UGC video ad featuring [image1], presenting product [image2].

15-second vertical 9:16 authentic UGC bakery sharing video, shot on a phone by a friend, bright natural daylight, casual handheld micro-shake, photorealistic, real skin texture, exact English lip-sync, no subtitles, no text overlays, no watermark.

Use the reference person: stylish young Black woman with dark skin, braided hair styled into two low buns with long face-framing braids, wearing a wide black hat, black sleeveless fitted top, white mini pleated skirt, white collar detail, black-and-white belt, soft confident expression. No handbag.

Background: a cosy modern bakery storefront, warm beige facade, a big glass window full of pastries, outdoor wooden tables and chairs, soft golden afternoon light, and an oversized croissant sculpture above the shop.

Hero food: an oversized giant croissant, almost half her body size, golden brown glossy crust, flaky layered ridges, buttery texture, crisp edges, warm bakery shine.

Shot 1: medium full-body outside the bakery. She holds the giant croissant with both arms, laughs at how huge it is, faces camera with mouth visible and says: "Okay, this is officially the biggest croissant I've ever held."
Shot 2: close-up of the croissant. She tilts it toward the camera to show the glossy crust, flaky layers and huge size, then says: "Look at these layers. It's literally bigger than my bag."
Shot 3: she takes one real bite from the edge, crunch sound, crumbs fall naturally. She lowers the croissant so her mouth is visible, reacts with widened eyes and says: "Wait. It's actually so buttery and crispy."
Shot 4: she hugs the giant croissant playfully, standing in front of the bakery window, and says: "I thought this was just for photos, but no, I'm eating this."
Final shot: full-body bakery exterior, giant croissant visible in her arms, soft smile to camera: "If you see this place, get the giant one."

Natural street ambience, bakery cafe sounds, crispy bite, light crumbs, live on-camera speech, exact lip-sync. The giant croissant stays flaky, golden and clearly visible throughout.

4 levers

Make it yours

What is safe to change. Most libraries publish only this list, which is why so many copied prompts come back worse than the original.

01
The line lengths
Eleven, ten, eight, twelve and nine words. Every line fits in about three seconds at conversational pace. This is the constraint the whole script is written around, not an accident.
02
The product
The oversized-object gag transfers to anything: a giant coffee, a giant burger, a comically large bottle. What has to survive is the size relationship — "almost half her body size" is a measurable instruction and "giant" is not.
03
The camera style
"Shot on a phone by a friend" and "casual handheld micro-shake" are two clauses doing the entire UGC look. Remove them and the same script produces a polished brand film, which is a different product.
04
The bite
Shot 3 is the only shot where something irreversible happens. UGC food content that never shows a bite reads as staged, and this is the one beat that cannot be moved to the end.

Three ways to break it

Hiding the mouth while a line is spoken
Shot 3 says "she lowers the croissant so her mouth is visible" — that clause exists because a giant pastry in front of a face is exactly where the model would otherwise put it, and lip-sync you cannot see reads as dubbing. Every shot with dialogue needs the mouth explicitly cleared.
Writing lines longer than about twelve words
Five shots in fifteen seconds gives each one about three seconds. At conversational pace that is eight to twelve words. A fourteen-word line does not get more time — it gets spoken faster, and the sync degrades first.
Dropping "no subtitles, no text overlays"
UGC is the format most heavily captioned in any training set, so this is where the model is most likely to burn text into the frame. On a fifteen-second five-shot video that is five chances to ruin the clip.

What it does

What the ugc oversized pastry template does

Five spoken lines in fifteen seconds is the most lip-sync work asked of the model anywhere in this library, and the prompt gets away with it by being unusually disciplined about one thing: it says, in every shot, where the mouth is.

Shot 1: "faces camera with mouth visible and says". Shot 3: "she lowers the croissant so her mouth is visible… and says". Shot 5: "soft smile to camera". That is not padding. A model handed a shot description involving a very large object held in both arms will frame the object, and the most natural place to put a croissant half the height of a person is directly in front of the person. The line still gets spoken; you just cannot see it being spoken, and the result reads as dubbed.

The word budget is the second discipline. The five lines are eleven, ten, eight, twelve and nine words. Nothing over twelve. Fifteen seconds across five shots is three seconds each, and three seconds of natural conversational delivery is eight to twelve words. Writing a fifteen-word line does not buy more time — the model compresses the delivery to fit, and compressed delivery is where sync fails first. Counting words before you write the shot list is unglamorous and it is the difference between this working and not.

The UGC register itself is carried by four clauses, all in the header: shot on a phone by a friend, casual handheld micro-shake, photorealistic real skin texture, and — critically — no subtitles. The first three produce the look. The fourth prevents the single most common artefact in this format, because short-form social video in any training set is overwhelmingly captioned, and a model asked for UGC will supply captions unless told not to.

"Real skin texture" is worth its own note. Video models default to a smoothed, retouched face, which is the exact opposite of the aesthetic that makes UGC persuasive. Two words move the output from advertisement to recommendation, which for this format is the entire conversion mechanism.

Shot 3 is the structural centre. It is the only shot in which something irreversible happens — the croissant gets smaller and does not get bigger again. UGC food content that never shows a bite reads as staged no matter how casual the camera is, because the implicit claim of the format is that the creator actually consumed the thing. Placing it third rather than last is also deliberate: it leaves two shots of reaction afterwards, and the reaction is the review.

Finally, the two reference images are doing different jobs and the prompt is explicit about which is which. Image one is the person, and the person is re-described in full in the prompt body — hair, hat, top, skirt, belt, no handbag. Image two is the product. Re-describing the person in text alongside the reference is not redundant here: across five shots the model needs the wardrobe as a checklist it can verify each shot against, and "no handbag" in particular is the kind of detail that quietly reappears by shot four.

Everything here runs in image to video.

6 questions

UGC oversized pastry — common questions

01
How many words of dialogue fit in a three-second shot?
Eight to twelve at natural pace. Longer lines are not given more time; they are delivered faster, and sync degrades before anything else does.
02
Why does the prompt keep saying the mouth is visible?
Because the model will otherwise frame the product in front of the face. Lip-sync you cannot see reads as dubbing, so every line needs its mouth explicitly cleared.
03
How do I stop captions being burned into a UGC clip?
Say "no subtitles, no text overlays" in the header. Short-form social video is the most heavily captioned material in any training set and the model reproduces that by default.
04
What makes the output look like UGC rather than an ad?
Four clauses: shot on a phone by a friend, handheld micro-shake, photorealistic real skin texture, natural daylight. The last one matters most — smoothed skin is what gives an ad away.
05
Should I re-describe the person if I supplied a reference photo?
Across several shots, yes. The written description acts as a per-shot checklist, and negatives like "no handbag" are exactly the details that drift back in by the fourth shot.
06
Why is the bite in shot three rather than at the end?
So there are two shots of reaction after it. The bite is the proof and the reaction is the review; ending on the bite gives you the proof with nothing to conclude from it.

More like this

Other templates that start from a still

Bring your own frame, or copy the image prompt into whichever model you already use.

All prompt ideas

Copy it, change three things, run it.

Every character is on this page. 15s at 2K costs $1.95.

Sign in free · 3 clips a day