全部, Twitter / X, Reddit, Instagram, MeiGen, prompts.chat, Wikiprompt
全部, Model-agnostic (7638), GPT Image 2 (5828), Midjourney (4830), Nano Banana (3820), GPT Image (3353), ChatGPT (3306), Nano Banana Pro (2744), Seedance 2.0 (1661), Nano Banana 2 (1008), Grok (587), Claude (573), GPT Image 1 (548), Grok Imagine (532), Gemini (514), Hailuo (462), Kling (448), Veo 3 (447), DALL-E (319), Seedance 2.5 (247), GPT Image 1.5 (246), Firefly (203), Z-Image Turbo (193), Runway (192), Seedream (189), Flux (181), Seedance (147), Leonardo (143), other (127), Sora (114), Veo (81)
全部, cinematic (16014), photorealistic (11027), editorial (5188), realistic (4878), minimalist (4369), fashion (3314), surreal (3274), dramatic (2845), portrait (2645), ethereal (1930), fantasy (1834), digital art (1768), vintage (1680), luxury (1587), illustration (1586), futuristic (1580), whimsical (1394), anime (1326), abstract (1325), hyper-realistic (1297)
全部, 单次 (20815), 创意生成 (14456), 模板 (14055), 系统提示词 (1267), 工作流 (598), 角色 (532), 智能体 (85)
关于 Wikiprompt, Blog, 统计, 创作者, API, MCP Server, Claude Skill, 使用条款, 隐私政策
4 个提示词

使用Project Neo进行场景建模、Adobe Firefly用于结构参考,以及第三方AI视频工具进行图像到视频生成的AI生成电影导演分步工作流程。


A three-step Midjourney workflow to generate complex scenes by starting with a simple gradient, using custom zoom with a new prompt, then varying for improved versions.
