OpenZhenZhen AI Prompt Gallery · MiniMax H3 · video

Colossus Breaks the City: Rooftop Escape

Colossus Breaks the City: Rooftop Escape — explore this MiniMax H3 creation on OpenZhenZhen, with its original prompt and a reusable generation example.

Colossus Breaks the City: Rooftop Escape — AI video creation
Colossus Breaks the City: Rooftop Escape

Prompt

Chinese prompt

创作一段15秒、16:9的超写实电影感奇幻动作短片,采用戏剧性的金色时刻光线,动作迅速清晰;不要文字、标志、喜剧元素或现代物品。0–2秒:从街道高度拍摄,一只巨大的石手冲破鹅卵石铺成的主街,震裂路面,扬起尘土与碎石,行人四散逃跑。2–5秒:跟拍一名成年男性战士和一名成年女性法师在瓦顶间奔跑,跃过一道狭窄的空隙;正在升起的石像巨人紧追在后。女性角色必须符合提供的身份参考图;参考图仅用于她的外貌,服装按此处描述保持不变。5–8秒:广角展现由高塔和桥梁构成的古老城市,巨大的石像巨人升至天际线之上,石质身躯上裂开橙色魔法光纹。8–10秒:在石桥上,法师转身朝巨人发射一道明亮的魔法冲击波,战士在她身旁直面威胁。10–12秒:冲击波在巨人身上迸发,尘土与碎石在桥边翻涌。12–15秒:两位英雄重重落在桥面上,随后惊恐地回望高耸于废墟城市之上的巨人。全片保持两位英雄的身份、面容和服装一致。

English prompt

Create a 15-second, 16:9 hyper-realistic cinematic fantasy action sequence with dramatic golden-hour lighting, fast readable movement, and no text, logos, comedy, or modern objects. 0–2s: From street level, a colossal stone hand bursts through the cobbled main street, cracking the pavement and throwing up dust and rubble as townspeople flee. 2–5s: Track an adult male warrior and an adult female mage as they sprint across tiled rooftops, leaping a narrow gap while the rising golem looms behind them. The female character must match the supplied identity reference image; use it for her appearance only, and keep her clothing as described here. 5–8s: A wide view reveals the ancient city of towers and bridges as the enormous golem rises above the skyline, its stone body split by glowing orange magical fissures. 8–10s: On a stone bridge, the mage turns and fires one bright magical blast toward the golem; the warrior faces the threat beside her. 10–12s: The blast flares against the golem as dust and debris billow around the bridge. 12–15s: The heroes land hard on the bridge, then look back in alarm at the fully risen, glowing golem looming over the ruined city. Keep the two heroes’ identities, faces, and outfits consistent throughout.

Topics and styles

View API model

Explore more AI image and video prompts · Try image generation · Try video generation · Create on the AI canvas