AI Video Prompt Examples You Can Copy and Adapt

"A lone red fox walks across fresh snow at dawn, breath visible in the cold air, slow tracking shot from the side, soft golden light, shallow depth of field, cinematic."

"Waves roll onto a black sand beach at sunset, foam catching the last light, slow pull-back to reveal the empty shore, warm tones, cinematic wide shot." Swap the animal, weather, or time of day to generate endless variations on the same reliable structure.

"A glass perfume bottle rotates slowly on a dark reflective surface, soft rim lighting catches the edges, slow push-in, minimalist, high-end commercial look."

"A pair of running shoes on a clean white pedestal, slow orbit around the product, bright studio light, shallow depth of field, crisp and modern." Replace the product and keep the lighting and motion to flatter almost any item.

"A young woman in a yellow raincoat stands in light rain, looking up slowly, soft overcast light, gentle push-in, shallow depth of field, cinematic portrait."

"An older man laughs in a warm cafe, steam rising from his coffee, handheld feel, soft window light, intimate close-up." Keep portrait motion subtle so the likeness holds, and lean on lighting for mood.

"A futuristic city at night, neon signs reflecting in wet streets, slow aerial descent between the towers, cyberpunk color, cinematic, moody lighting."

"Glowing particles drift through dark space and slowly swirl into a spiral, deep blues and purples, slow camera rotation, abstract, dreamlike." These genres reward bold lighting words and a single sweeping move.

"Melted chocolate drips slowly over a fresh strawberry, soft warm light, extreme close-up, slow push-in, glossy and appetizing."

"Steam rises from a bowl of ramen on a wooden table, chopsticks lift noodles, warm overhead light, shallow depth of field, cozy and inviting." Food clips thrive on tight framing, slow motion cues, and warm light.

Keep the skeleton and change the content. Each example names one subject, one action, one setting, one camera move, and one lighting cue. Replace any of those parts, but do not add a second camera move or a second main action, which is where clips usually break down.

When a result is close but not right, change a single element and regenerate. Adjust the lighting word, slow the camera, or simplify the action. This deliberate, one-variable-at-a-time approach turns these examples into a reliable toolkit rather than a lucky guess.