A 15.7s portrait AI video based on this source image. Movement and details may vary from the image.
Generation direction: portrait. The source image is the reference for the matching AI video; details and movement may vary.
Image and video details
Video length
15.7s
Video format
Portrait · 768 × 1024
Source image
880 × 1184
Published
Free viewing and creation costs
All platform-published images and videos are free to browse and watch. No account, credits or daily viewing allowance is required. Replay, switch videos and return as often as you like. Public work pages include available generation prompts and source themes. Creating new images or videos, chatting and voice features still use credits at the displayed price. Creation prices.
This instruction describes the intended video; details and movement in the finished clip may vary.
subject_definitions: <Picture 1> is the appearance reference for the female lead only; it is not treated as a locked first frame. <Subject 1> is the adult woman identified with <Picture 1>, described in the source text as a tanned redhead with a slim figure, freckles, green eyes, glasses, and an attractive, well-proportioned look. She is the only on-screen person. Her wardrobe is stylish, sexy, and curve-accentuating while remaining non-nude: a fitted cocktail dress worn over a bra and panties. Features called out to avoid (hat, bunny ears, cat ears, green skin, extra limbs, deformed anatomy, visible sexual organs) are not part of this subject.
summary: A 13-second 3:4 hyperrealistic short of <Subject 1> from <Picture 1> greeting the viewer, dancing as if on a private date, and inviting them closer. The still defines her identity; the video adds lounge lighting, stylish non-nude attire, original flirt dialogue, and fast energetic music. No other faces appear.
retention_analysis: <Subject 1> identity from <Picture 1> is fully_preserved: adult tanned redhead, slim body, freckles, green eyes, glasses, attractive facial structure, sole on-screen character. Clothing is partially_preserved versus the collected nude tag because the brief forbids nudity and sexual organs; she instead wears a sexy fitted dress over required undergarments. Pose is partially_preserved: the custom still pose becomes greeting, dance, and lean-in toward camera. Setting is weak_reference because the source still does not specify a location; a dim lounge is added to support a date-and-dance story. Negative traits (hat, animal ears, extra limbs, deformed face or hands, genital depiction, statue stiffness, green skin) are fully_preserved as absences.
detailed_description: [Shot 1] Hyperrealistic portrait in a 3:4 frame, intense surface detail, precise and clean rather than painterly. Medium shot of <Subject 1> from <Picture 1> standing in the foreground of a dim modern lounge: dark seating, a low table, and warm practical lamps behind her. Soft key light from camera-left sculpts her tanned face and slim figure; cooler rim light separates her from the background. She wears stylish, sexy, curve-accentuating date attire: a fitted dark cocktail dress over a bra and panties, no nudity, no sexual organs, no hat, no bunny ears, no cat ears. Glasses sit on her nose; red hair frames freckles and green eyes as she looks into the lens. No other people and no other faces. She steps in from a slight pause, hips shifting, then lifts one hand toward the camera as if greeting a date, a teasing smile forming. (S1) says, eyes on the viewer, <d>[English] There you are. I dressed up just for you.</d>
[Shot 2] At 00:04.300, The camera cuts to a slightly lower medium-close shot of the same <Subject 1>, still the only person in frame. Warm key light remains camera-left on her dress fabric and glasses. She immediately rolls her shoulders and dances toward the lens with tight, energetic steps, one hand tracing her waist, the other reaching out as if pulling the viewer onto the floor. The dress stays in place and continues to cover her; undergarments remain on, with no nudity and no depiction of sexual organs. Her mouth opens on a teasing breath, green eyes locked on camera, freckles and tan skin catching highlights. No extra limbs, no deformed hands, no second face. (S1) says while dancing, <d>[English] Dance with me. Don't look away.</d>
[Shot 3] At 00:08.600, The camera cuts in to a close three-quarter portrait of <Subject 1> from <Picture 1>. She leans toward the lens, glasses catching a small highlight, red hair falling beside her freckled cheek. The lounge lamps stay out of focus behind her; the same camera-left key keeps her face and collarbone separated from the dark background. After the dance she steadies, tucks hair behind one ear, then rests her fingers near her collarbone and holds a slow, inviting smile. She never removes the dress, never appears nude, and no sexual organs are shown. No hat, no statue-like stiffness, no other individuals. (S1) says softly to the viewer, lips matching the line, <d>[English] Tonight it's just us. Come closer.</d>
overall_soundscape: A quiet lounge bed of muffled room tone and distant glass clinks sits under the picture, never crowding the voice. Heels tap on a hard floor, dress fabric rustles, and her breath catches as she steps in and dances. Non-verbal voice presence is close-miked: a smile in the inhale, a teasing exhale between lines, then a softer close-mic closeness on the final lean-in.
non_diegetic_music: A fast, driving electronic-pop cue with a tight four-on-the-floor kick, bright synth stabs, and a rising bass line. Energy stays high from the first beat, with a short lift into the dance cut and a warmer, still-rhythmic close under the last invitation.
The gallery pairs collected source images with completed AI-generated clips. The image is a reference for the video, so movement and details may vary. Source titles are preserved where available. Otherwise, a short title uses neutral visual terms from the generation prompt or source themes, with a reference number identifying the study. Each work page shares its available video generation prompt and source theme tags. The prompt describes generation intent, including setting, lighting and framing. It is not image recognition or a guarantee about every video frame.