FLUX 3 Max UGC Prompts: 5 Fill-In-the-Blank Templates

FLUX 3 Max UGC prompts work best as fill-in-the-blank templates: one creator, one action, handheld vertical framing, natural light, and a short spoken line with room tone and no music. Below are five 9:16 templates (testimonial, unboxing, product in hand, problem-solution and street style), each with a filled example, its duration and its credit cost.

Every clip FLUX 3 Max generates on VIDEO AI ME comes back with sound made in the same pass, so a UGC prompt directs the voice and the room as well as the picture. The templates follow the four-part order from the FLUX 3 Max prompt guide: subject and action, camera, light and look, then sound. They are teaching material: fill them in for your product and judge every result yourself.

What makes FLUX 3 Max UGC prompts work?

UGC works because it looks unproduced. Each part of the prompt should protect that:

• Subject and action: a specific person (age range, clothing) doing one ordinary thing with the product in a real room. One beat per clip.
• Camera: "handheld selfie framing, eye level, slight natural sway" or a phone "propped on the counter". Avoid gimbal moves and orbits, which read as studio work.
• Light: window light, a desk lamp, overcast daylight. Name the source.
• Sound: the line in double quotes with the speaker named, the room tone, the sound of the action, then "no music". A polished music bed breaks the format faster than anything in the picture, and TikTok's creative best practices call sound vital to the TikTok experience, so direct it rather than leave it to chance.

Two rules matter here. Black Forest Labs' prompting documentation for FLUX 3 notes that a quoted line without a visible speaker or a voiceover cue may be treated as text that belongs in the frame, so always write "she says" or "a woman's voiceover says". And name the sound source you want rather than asking for silence: "quiet kitchen tone, a fridge hum" instead of "silent".

Keep captions, prices and offers out of the prompt. Add them in the editor, where they stay sharp and editable.

How do you fill in a UGC prompt template?

Each template uses square brackets for the parts you replace. Fill every bracket, delete the brackets, and keep the order of the sentences.

The bracket that fails most often is [LINE]. Budget roughly two to three spoken words per second of clip, then write fewer:

FLUX 3 Max costs 17 credits per second at 720p and 29 credits per second at 1080p. Set every template to 9:16 in the editor, which matches TikTok's in-feed ad specifications, where vertical 9:16 is the recommended ratio.

Which FLUX 3 Max UGC prompts should you start with?

Template 1: Testimonial to camera

[CREATOR: age, look, clothing] sits [PLACE] holding [PRODUCT: colour,
material, size] and talks to the phone camera, [SMALL GESTURE]. Handheld
selfie framing, eye level, slight natural sway. [LIGHT SOURCE AND
DIRECTION], natural skin tones. [She/He] says, "[LINE]." [ROOM TONE],
[ONE BACKGROUND SOUND], no music.

Filled example, 9:16, 10 seconds, 720p, 170 credits:

A woman in her early thirties in a grey sweatshirt sits cross-legged on a
sofa holding a small amber bottle of hair oil and talks to the phone camera,
turning the bottle in her hand. Handheld selfie framing, eye level, slight
natural sway. Soft daylight from a window on the left, natural skin tones.
She says, "Two drops on the ends every night. That is the whole routine."
Quiet living room tone, a dog barking far off, no music.

Template 2: Unboxing

Two hands open [BOX] on [SURFACE], fold back [INSERT] and lift out
[PRODUCT], then hold it up toward the camera. Top-down, handheld, slight
natural movement. [LIGHT]. A [woman's/man's] voiceover says, "[LINE]."
[PACKAGING SOUNDS], [ROOM TONE], no music.

Filled example, 9:16, 12 seconds, 720p, 204 credits:

Two hands open a matte white mailer box on a light oak desk, fold back pale
pink tissue paper and lift out a sage green ceramic mug, then hold it up
toward the camera. Top-down, handheld, slight natural movement. Soft daylight
from a window above the desk, clean natural grade. A woman's voiceover says,
"Okay, the colour is even better in person. And it is heavier than I
thought." The tear of tape, crinkling tissue paper, quiet room tone, no music.

The speaker is behind the camera, so the line is marked as voiceover.

Template 3: Product in hand

Close-up of a hand holding [PRODUCT] [PLACE], [ONE DEMO ACTION]. Handheld
close-up, slight push-in. [LIGHT]. A [woman's/man's] voiceover says,
"[LINE]." [THE SOUND OF THE ACTION, tied to the moment it happens], [ROOM
TONE], no music.

Filled example, 9:16, 8 seconds, 720p, 136 credits:

Close-up of a hand holding a frosted glass pump bottle of hand cream over a
bathroom sink, pressing the pump twice onto the other palm and rubbing the
cream in. Handheld close-up, slight push-in. Warm light from a mirror lamp
above, soft natural grade. A man's voiceover says, "No greasy feel. It is
gone in ten seconds." The pump clicks twice as the thumb presses down, a
soft rub of hands, quiet bathroom tone, no music.

The sound is written against the action ("the pump clicks twice as the thumb presses down"), which gives the effect a moment to land on.

Template 4: Problem-solution in two clips

One clip, one beat. A problem and its solution are two beats, so this template is two prompts joined in the editor. Write the creator and the place identically in both.

Clip A: [CREATOR] [STRUGGLES WITH THE PROBLEM] in [PLACE], then looks at
the phone camera, [REACTION]. Handheld, eye level, slight natural sway.
[LIGHT]. [She/He] says, "[PROBLEM LINE]." [SOUND OF THE PROBLEM], [ROOM
TONE], no music.

Clip B: [SAME CREATOR] in [SAME PLACE] [USES THE PRODUCT], then [REACTION
TO CAMERA]. Handheld, eye level, slight natural sway. [SAME LIGHT].
[She/He] says, "[SOLUTION LINE]." [SOUND OF THE PRODUCT], [ROOM TONE], no
music.

Filled example, two clips at 9:16, 8 seconds each, 720p, 272 credits for the pair:

A man in his late twenties in a white t-shirt tugs at a tangled knot of
charging cables on a cluttered desk, then looks up at the phone camera,
exasperated. Handheld, eye level, slight natural sway. Flat overhead office
light, neutral grade. He says, "Every single morning. Three cables, one
knot." Cables scraping on the desk, quiet room tone, no music.

A man in his late twenties in a white t-shirt sets his phone, earbuds and
watch on a slim charging pad on a tidy desk, then leans back and nods at the
phone camera. Handheld, eye level, slight natural sway. Flat overhead office
light, neutral grade. He says, "One pad. Everything charges. The desk is
mine again." A soft click as each device lands on the pad, quiet room tone,
no music.

Two separate text to video generations are not guaranteed to show the same person. If continuity matters, start both clips from the same photo of the creator in image to video, covered below.

Template 5: Street style

[CREATOR AND OUTFIT, including the PRODUCT] walks toward the camera along
[STREET], [SMALL ACTION WITH THE PRODUCT], then stops and [GESTURE].
Handheld, leading [her/him] at chest height. [LIGHT]. [She/He] says,
"[LINE]." [STREET SOUNDS WITH PLACEMENT], [FOOTSTEPS], [MUSIC LOW IN THE
MIX or no music].

Filled example, 9:16, 10 seconds, 1080p, 290 credits:

A woman in her mid twenties in an oversized black blazer and wide-leg jeans
walks toward the camera along a sunny shopping street, adjusting the strap
of a tan leather crossbody bag, then stops and spins once. Handheld, leading
her at chest height. Bright midday sun, crisp contrast, warm grade. She
says, "It fits my phone, my keys and my whole personality." Street chatter
on the left, a passing scooter on the right, heels on paving stones, light
upbeat music low in the mix.

Street style is the one UGC format where light music can fit, because outfit content often carries a track. Placement cues ("chatter on the left, a scooter on the right") keep the street from sounding like a flat wash.

How do you adapt FLUX 3 Max UGC prompts for image to video?

Attach one photo and it becomes the first frame. The photo now carries the creator, the product and the room, so the prompt only needs motion and sound:

[She/He] lifts [the product] toward the camera, [SMALL ACTION], and says,
"[LINE]." The camera stays handheld at eye level with a slight sway.
[ROOM TONE], [SOUND OF THE ACTION], no music.

Use a photo you have the rights to, with the person's consent, and leave the ratio on auto if the photo is already vertical. FLUX 3 Max takes one reference image. If a brief needs several, Seedance 2.5 and MiniMax H3 Max accept multiple reference images. For the wider UGC workflow on VIDEO AI ME, see the AI UGC generator.

What should the spoken line say?

The line is an ad claim, whoever appears to say it. Keep it to what the product actually does, and avoid invented results ("I lost ten pounds") and invented histories ("I have used this for five years").

The FTC's Endorsement Guides Q&A says endorsements must reflect the honest opinions of the endorser and that a material connection should be disclosed clearly and conspicuously. Read it, and the AI UGC ads FTC disclosure guide, before a campaign runs. For writing the line itself, the AI UGC ad script guide covers hooks and structure.

What does a batch of UGC variations cost?

One run of each template at 720p (templates 1, 2, 3 and 5 as single clips, template 4 as its pair) is 170 + 204 + 136 + 272 + 170 = 952 credits. The Starter plan's 1,400 monthly credits cover that batch with room for a 1080p delivery of the winner, which at 10 seconds is 290 credits.

If you test several hooks per template every week, the Pro plan's 5,600 monthly credits cover about 32 ten-second clips at 720p. Compare allowances on the pricing page, and see how these templates fit a full campaign in FLUX 3 Max for UGC ads.

Frequently Asked Questions

How do you write a UGC prompt for AI video?

Describe one specific person doing one ordinary thing with the product in a real room, then add handheld eye-level framing, a named natural light source, and the sound: a short line in double quotes with the speaker named, the room tone, the sound of the action and no music.

What aspect ratio should UGC ads be?

Vertical 9:16 for short-form feeds. TikTok recommends 9:16 for in-feed ads. In FLUX 3 Max you set the ratio in the editor, not in the prompt.

How long should a UGC video ad clip be?

A single-beat UGC clip usually needs 8 to 12 seconds. FLUX 3 Max generates up to 20 seconds in one clip, and longer ads are built by joining several clips in the editor, one beat each.

Should AI UGC videos have music?

Usually not for talking or demo clips, where a polished music bed breaks the creator feel. Write no music explicitly, because FLUX 3 Max always generates audio and will otherwise decide for you. Street style and outfit content can carry light music low in the mix.

Do AI UGC ads need a disclosure?

Check the FTC's Endorsement Guides before you run them. They say endorsements must reflect honest opinions and that a material connection should be disclosed clearly and conspicuously. Keep every spoken claim to something your product actually does.

Can FLUX 3 Max turn my product photo into a UGC video?

Yes. Attach one photo and it becomes the first frame of an image to video clip, with audio still generated. Your prompt then describes only the motion and the sound. For briefs that need several reference images, Seedance 2.5 and MiniMax H3 Max accept multiple images.