Gem Nana AI Prompt Library
The legacy image, video, and prompt catalog is now inside the new SaaS foundation with search, tags, and detail pages. Favorites, memberships, and gated downloads can build on this next.
Showing 34 results
1 / 2
Seedance 2.0 用参考图驱动舞蹈动画
用运动参考图结合 Seedance 2.0，并用 GPT Image 2.0 生成分镜姿态，快速做舞蹈动画。
四组背景提示词（Part3）
提供星辰、深渊、峡谷与新绿四类奇幻背景提示词，可直接使用或再创作。
GPT Image 2 数学可视化提示词模板
提供可复用的GPT Image 2数学信息图提示词模板，强调直观结构与学术排版风格。适合生成清晰不杂乱的讲义式海报。
偶像风服装提示词模板
提供“偶像风”西装背心裙的详细提示词，含配色、结构与细节描述，便于生成一致造型。
东方史诗海报自动叙事生成
用轮廓剪影将角色世界观扩展成电影级叙事海报，含光影质感与空间层次设计。
GPT-Image 2.0稳定可用
GPT-Image 2.0已稳定上线，可生成视觉精美的奇幻食谱图片与步骤。
Seedance 2.0 生成一生电影镜头
用Seedance 2.0打造15秒16:9电影级连续横版卷轴：全程一镜到底、环境无缝变换。医院到老年，音效与情绪逐段匹配。
Gemini Nano Banana Pro人像抓拍提示词
提供iPhone照片级真实自动扶梯女性抓拍提示词，强调身份一致、自然光与无滤镜。
Seedance 2.0虹膜镜头提示解析
用Seedance 2.0实现“虹膜快推+镜面破碎世界”的镜头效果，适合测试随机语言提示解读。
服装提示分发与改造技巧
可自由分发服装提示，并根据喜好改色、增删元素。文中也提供男角色佩戴的提示与技巧。
双重曝光剪影电影海报提示
用剪影轮廓+世界构建打造收藏版叙事海报，并加入边缘羽化与纸纹质感。
Gemini Nano Banana Pro写实街拍提示
描绘昏暗停车场年轻女子偷拍感肖像，含姿态、光线与强运动模糊细节。
Seedance 2.0：魔镜
Seedance 2.0 面向影视广告行业，支持多模态输入并进行视频生成与编辑，支持网页端与API按次计费。
Seedance 2.0 网格编排技巧
用Grok生成技巧清单，再用网格与banana控制动作自由度。适合SD2编排滑冰动作与剪辑。
Gemini Nano Banana 2写实提示词
用Gemini Nano Banana 2生成超写实巨型可乐罐山景，含探险者与冷色电影光效。提示覆盖构图、质感与8K细节。
可爱3D渲染提示词分享
提供“揉捏粘土图标风”可爱3D提示词模板，包含材质、配色与光影要求。适合快速生成高分辨率、抛光玩具质感作品。
地下室独自拍摄的15秒野生动物追逐镜头设计
揭示单人制作野生动物纪录片时，如何实现15秒超真实连续追逐、尾巴抓持与自然光线的无剪辑镜头要点，并探讨使用 Kling 3.0 等工具的实现思路。
Long-exposure street portrait: the contrast between a silent woman and a ginger cat in the dusk square.
In a twilight city square, a young woman and a ginger cat stand side by side in silence, using long exposure and background blur to create contrast, conveying a quiet and deeply affectionate mood. The image emphasizes the tactile quality of fabrics, warm color accents, and the artistic effect of ultra-high resolution.
Winter Village Portrait: The Gaze of the Shawl-Wrapped Snowflake Girl and the Tender, Silent Light and Shadow
In a tranquil winter countryside scene, a young woman turns back and gazes, with snowflakes clinging to her shawl, highlighting a natural texture and a warm contrast. The shot employs soft, diffused light and shallow depth of field to convey a cinematic feel.
Surreal-style prompts for ingredient illustrations for the next step in designing an oil-brand.
Prompts for surreal, high-end commercial illustrations generated with ChatGPT Image 1.5, emphasizing embedding the product bottle into a natural landscape made from core ingredients, and providing ideas for the next-step ingredients.
Convert the original photo into four images in dramatic, high-contrast styles.
Convert the original photo into four dramatic styles, using extreme wide-angle and dynamic angles to render cinematic lighting, high-definition contrast, and delicate textures in ultra-realistic imagery.
Celebrity face recognition tagging and zoomed-in focus effects in city surveillance footage.
Celebrity portrait generation prompt set against an urban surveillance backdrop, including AI face-recognition tags, amplified focus, and a low-saturation cool-toned style, depicting multiple stars in dynamic scenes on a busy street.
Vertical Chinese retro educational infographic poster design: one-click packaging, ready to use out of the box.
This tweet describes a design specification for a vertical, traditional Chinese retro-style educational infographic poster, including modules such as core theme, classic quotes, six core viewpoints, a relationship map, a timeline and character cards, and other components. It emphasizes unified typography, white space ratio, high-resolution output, and bidirectional vertical/horizontal adaptability, making it suitable for large-format display and exhibition use.
One face, six emotions: AI-generated collage artworks of women in six moods.
This tweet introduces using Google's Gemini AI Nano Banana 3 Pro tool to create a six-panel collage of a 3D comic-style female portrait, displaying six different emotions: happy, surprised, serious, cute, playful, and confident, blending high-end digital art with rich colors.