这是一套三步式的电商详情页视觉设计提示词。第一步根据产品参考图联网搜索并反推产品专属的电商视觉风格体系，按整体调性、画面质感、布光逻辑、色彩体系、字体规范、构图规范六大模块输出，强调场景共情、卖点可视化与移动端竖屏适配。第二步提取产品成分、使用说明、保质期等信息，识别材质、结构与包装特色，并确认是否需要模特、场景和数据可视化。第三步以一线电商视觉设计师身份，依据前两步报告生成 14 屏 2:3 竖版详情图提示词，每屏包含中文提示词、负面提示词、产品卖点、字体规范与构图规范，并补充字体样式字号一致性、整体配色统一和免费商用字体标注等要求。

中文版提示词

第一步：请根据我上传的产品参考图，联网搜索该产品相关品类的详情风格定位，依据产品属性和品类属性，反推出这套产品专属的电商详情【视觉风格体系】。严格按照固定六大模块输出：整体调性、画面质感、布光逻辑、色彩体系、字体规范、构图规范，风格定位贴合电商爆款逻辑，避免纯色背景，主打场景共情、卖点可视化、移动端竖屏适配，文案格式统一字体类型，字号/字重层级分明，行距舒适，适配手机竖屏阅读。

第二步：提取成分信息、配料表、营养成分；提取使用说明、注意事项；提取生产日期、保质期、储存方式。产品细节识别：识别产品材质质感（光滑/粗糙/哑光/高光等）；识别产品结构特点（可拆卸/折叠/便携等）；识别包装特色（自封袋/泵头/喷雾/滴管等）。特殊需求（可选）：【是否需要模特】是/否（如需要，请描述模特类型）；【是否需要场景】是/否（如需要，请描述场景类型）；【是否需要数据可视化】是/否；【其他特殊要求】：[如：必须包含产品实物、需要对比图、需要用户评价等]。

第三步：任务说明：我想为我的产品设计一套电商详情图，帮我生成电商详情图设计系统提示词（共 14 屏，2:3 竖版）。角色设定：你是国内一线全品类全平台电商视觉全案 AI 设计师，精通淘宝、京东、拼多多、抖音电商等主流平台转化型视觉设计，具备商业摄影导演、信息可视化设计、中文爆款广告文案策划的全链路能力，能为目标产品生成全新的专属场景、色调和光影。请根据第一步的【视觉风格体系】和第二步的【产品信息智能提取报告】生成极致详细的提示词。视觉要求：1. 使用上述体系与报告推荐的视觉风格、文字效果、字号/字重层级；2. 视觉大胆新颖、场景高清、设计高级统一、无割裂感；3. 突出产品核心卖点，卖点可视化，展示角度多元化；4. 标题文字在画面顶部，采用左对齐/右对齐/居中对齐；5. 部分页面可不展示产品，以产品功能或材料进行可视化呈现。生成规范：1. 每一屏都输出中文提示词及负面提示词；2. 每条提示词包含 2:3 竖屏比例；3. 每条提示词包含产品卖点、字体规范（层级分明的字体体系、字号/字重层级）、构图规范（竖屏适配的模块化场景构图）；4. 将每一屏内容分别整理成一段文字以便直接拿去 AI 平台出图，每屏详细包含视觉要求、主标题和副标题、信息布局、排版形式、设计感强调，内容越详细越好。额外关键提示词：1. 保持字体样式、字号大小的一致性；2. 保持整体风格与详情页配色一致，整体色调不要有太大区别；3. 标注所使用的字体名称，且必须为可免费商用的字体（根据详情页风格体系调整）。请阅读以上分析报告后立即执行提示词生成。

英文版提示词

Step 1: Based on the product reference image I upload, search the web for the detail-page style positioning of this product category. Using the product's attributes and category characteristics, reverse-engineer a dedicated e-commerce detail-page [visual style system] for this product. Output strictly in six fixed modules: overall tone, visual texture, lighting logic, color system, typography standards, and composition standards. The style should fit viral e-commerce logic, avoid solid-color backgrounds, emphasize scenario empathy, visualize selling points, and adapt to mobile portrait reading. For copy, unify font types, make font size/weight hierarchy clear, and keep comfortable line spacing for portrait mobile reading.

Step 2: Extract ingredient information, ingredient list, and nutrition facts; extract usage instructions and precautions; extract production date, shelf life, and storage method. Product detail recognition: identify material texture (smooth/rough/matte/glossy, etc.); identify structural features (detachable/foldable/portable, etc.); identify packaging features (zipper bag/pump/spray/dropper, etc.). Special requirements (optional): [Need model] yes/no (if yes, describe the model type); [Need scene] yes/no (if yes, describe the scene type); [Need data visualization] yes/no; [Other special requirements]: [e.g., must include the physical product, need comparison images, need user reviews, etc.].

Step 3: Task: I want to design a set of e-commerce detail images for my product. Help me generate a design-system prompt for e-commerce detail images (14 screens in total, 2:3 portrait). Role: You are a top-tier, full-category, all-platform e-commerce visual designer in China, proficient in conversion-oriented visual design for Taobao, JD.com, Pinduoduo, and Douyin e-commerce, with end-to-end capabilities in commercial photography direction, information visualization design, and Chinese viral ad copy planning. You can generate brand-new exclusive scenes, tones, and lighting for the target product. Based on the [visual style system] from Step 1 and the [product information extraction report] from Step 2, generate extremely detailed prompts. Visual requirements: 1. Use the visual style, text effects, and font size/weight hierarchy recommended by the system and report; 2. Make visuals bold and novel, scenes high-definition, and the overall design aesthetic, premium, and consistent with no sense of fragmentation; 3. Highlight the product's core selling points, visualize selling points, and diversify display angles; 4. Place title text at the top of the image using left/right/center alignment; 5. Some pages may omit the product and instead visualize its functions or materials. Generation rules: 1. Output both a Chinese prompt and a negative prompt for each screen; 2. Include a 2:3 portrait ratio in every prompt; 3. Include product selling points, typography standards (a clear font hierarchy with distinct size/weight levels), and composition standards (modular scene composition adapted to portrait screens) in every prompt; 4. Organize each screen into a single paragraph so I can paste it directly into an AI image platform, and include visual requirements, main title and subtitle, information layout, composition form, and design emphasis in as much detail as possible. Additional key prompts: 1. Keep font style and size consistent; 2. Keep the overall style consistent with the detail-page color scheme and avoid large tonal differences; 3. Specify the font names used, and they must be free for commercial use (adjusted according to the detail-page style system). After reading the analysis reports above, immediately execute prompt generation.

🛠️ **适用 AI 工具**：Midjourney、Stable Diffusion、即梦、豆包、DALL·E、Flux