Free Vidu AI Video Generator: Image & Text to Video Online

Use the provided start frame as the first shot: a rainy nighttime city street with glossy reflections, a large glowing Christmas tree in the background, pedestrians with umbrellas passing behind; the boy (left) and girl (right, wearing a knit beanie) face each other in the foreground—camera in a medium close-up at eye level, shallow depth of field, cinematic warm bokeh from the tree and streetlights, gentle handheld micro-movement; light rain falls steadily with visible droplets and subtle mist, water beads on clothing and hair; the girl slowly steps closer, lifts her chin, and reaches up to rest a hand lightly on the boy’s chest/shoulder as he softens his expression and leans in; they close the distance and share a tender kiss, holding for a moment while umbrellas drift past in soft focus; keep their faces consistent with the reference image, maintain realistic skin texture, natural motion, and romantic, cozy holiday mood; 24fps, 4K, filmic color grade, no text, no logos, no flicker.

An Eastern young female dancer with long hair, wearing a light, sheer, flowing traditional-style long dress with long sleeves, holding a short knife/dance blade. In an open classical stage setting or in front of a clean background, the camera follows in a medium shot. She first makes a light jump into the air; with both arms, she guides the blade in a smooth arc, and her body completes one full 360° rotation in midair, with the movement continuous and steady. She then lands smoothly and finishes by drawing in and holding a final pose—elegant and rich in dance artistry. A gentle breeze lifts and visibly ripples the skirt and long sleeves. Soft, cinematic lighting. Clean, stable footage with no shaking, no deformed fingers, and no weapon distortion. Overall: an aesthetically beautiful, realistic style.

Old photo, photo-to-video effect in black-and-white: start from a high, top-down overhead view like an aged vintage photograph, showing a young woman seated on a large ornate patterned rug in a dim wooden room, wearing an off-shoulder corseted gown with puff sleeves and a full skirt spread around her, hair pinned up neatly and classic makeup; animate the still with subtle film grain, gentle flicker, and soft contrast, then the camera slowly pans/glides downward from above toward her face as she looks up into the lens with a calm, intense expression, and as the camera settles closer she slightly softens her gaze and gives a small warm smile directly at the camera, maintaining the authentic "old photo brought to life" feel.

POV shot:I am lounging on the sofa when my girlfriend sits close to me. She is wearing an oversized white button-down shirt, slightly unbuttoned. Her hair is damp and messy from the shower. She leans in with a playful grin, reaching out as if trying to block or grab the camera. High quality, photorealistic, authentic look.

A cinematic transformation shot starting with a man in a black coat and white turtleneck standing stoically in front of a tall, modern glass skyscraper under a blue sky; suddenly, a massive, fiery explosion erupts from the building behind him, filling the background with billowing orange flames and smoke while the man remains completely calm and unmoving, as the glass facade shatters and is consumed by the inferno in his back.

Apply a magical clown makeup effect to the same woman's face: track her facial features frame-by-frame and smoothly paint on a theatrical clown look that stays perfectly aligned during head turns and expressions—clean white base across the face, bright red round nose, vivid red lips extended into a soft upturned smile, and bold blue star/diamond shapes around both eyes—while preserving her identity, skin texture, pores, natural shadows, and the original lighting; keep hair, background, and camera style unchanged, maintain natural blinking and mouth movement, and ensure seamless edges at the hairline/jawline with zero flicker, warping, or ghosting.