PROMPT RADAR / VIDEO视频叙事
🌆 I just directed my own Blade Runner-style cyberpunk masterpiece on @l
🌆 I just directed my own Blade Runner-style cyberpunk masterpiece on @lilidiai! Workflow: → Powerful prompt: "futuristic cyberpunk city at dusk, neon reflections, cinematic lighting" → Generated 6 breathtaking variation
@ech0_speaks 1.0k 热度 2.0 万 浏览
编辑先看
它为什么值得进入灵感库
这条内容值得留下的，不是表面的风格词，而是它提供了一条可以继续实验的画面路径。 把一段描述拆成起始状态、动作变化、镜头响应和结束画面，连续性比形容词更重要。
01 / 原始方法
动作与镜头 Prompt
🌆 I just directed my own Blade Runner-style cyberpunk masterpiece on @lilidiai! Workflow: → Powerful prompt: "futuristic cyberpunk city at dusk, neon reflections, cinematic lighting" → Generated 6 breathtaking variations using GPT Image 2 → Transformed them into a smooth, ultra-cinematic video with Kling 3.0 Flying cars gliding through rain-soaked neon streets, towering holograms, dramatic volumetric lighting, and that signature moody atmosphere—pure Hollywood sci-fi magic in minutes Everything in one sleek workspace: Flux, GPT Image 2, Kling 3.0, Sora 2 & more. No tool switching. Just pure creation. Level up your AI filmmaking → #BladeRunner #Kling3 #AIVideo #Cyberpunk #GPTImage2 #LilidiAI #CinematicAI #ImageToVideo
02 / 复刻路径
不要一次改完所有变量
用 GPT Image 2 复刻时，建议先单独生成稳定首帧，再把动作拆成“开始、变化、收束”三段；每段只保留一个主要运动。
- 主体：🌆 I just directed my own Blade Runner-style cyberpunk masterpiece on @lilidiai! Workflow: → Powerful prompt: "futuristic cyberpunk city at dusk, neon reflections, cinematic lighting" → Generated 6 breathtaking variation
- 视觉方向：实时 AI 视频案例；查看原帖媒体与作者说明
- 镜头与构图：按原帖镜头流程执行；查看原帖媒体与工作流
- 交付约束：auto，保留原作者方法与归属；复刻时避免水印、低清、畸变、主体漂移和无关文字。
负向约束
保留原作者方法与归属；复刻时避免水印、低清、畸变、主体漂移和无关文字。
模型建议
GPT Image 2 / Kling / Seedance / Veo / Flux。先用原始 Prompt 建立基准结果，再根据模型对自然语言、镜头运动或文字排版的能力做局部改写。
来源与编辑原则