FastArt.ai
⌕
30,883 CURATED PROMPTS

Prompt Library

All · 30,883Images · 24,022Videos · 6,861
#portrait#character#poster#fashion#Seedance#fantasy
24,022 results · Page 155 of 1001
IMAGE · chatgpt

AI全自动超长信息图生成器

Please automatically create an ultra-long infographic based on the topic input by the user. This is not an ordinary image generation task. Please treat this task as a complete long infographic production project, automatically completed in this round of tasks: data understanding / web verification when necessary / information architecture planning / master canvas design / segmented continuous generation / seamless stitching / error checking / final export. Do not simply compress extreme ratios into ordinary vertical images, and do not just generate a single short image pretending to be a long image. Do not show intermediate processes, drafts, segmented images, or analysis processes; only return the final complete product. 【User Input Variables】 Topic: {topic} Final Width: {width} Final Height: {height} Final Aspect Ratio: {aspect_ratio} Visual Style: {style} Information Density: {density} Output Language: {language} Web Access: {web_mode} Text Generation Mode: {text_mode} Color Palette: {palette} Main Visual Anchor: {visual_anchor} Output Format: {format} 【Automatic Processing Rules】 1. Automatically determine the category the topic belongs to, such as: History / Technology / Business / Geography / Tutorial / Figure / Culture / Product / Industry Analysis / Trend Summary. 2. Automatically design the most suitable information structure according to the topic type, without using rigid templates. 3. If the topic depends on real facts, timelines, data, development history, phase divisions, representative figures, real cases, or the latest information, automatically verify on the web and extract key information. 4. If the topic is conceptual, aesthetic, knowledge organization, or non-time-sensitive content, the structural design can be completed directly based on existing knowledge. 5. Automatically generate a main visual anchor matching the topic, such as: rivers, paths, timelines, energy lines, knowledge trees, map veins, city roads, star maps, process flows, product chains, etc., and make it run through the entire long infographic. 【Content Structure Requirements】 Automatically generate a reasonable information structure based on the topic. Include at least: - Main Title - Subtitle / Introduction - 4 to 10 main sections - Key points under each section - Representative visual elements - Bottom summary / Enlightenment / Conclusion / Method checklist If the topic is History, it should prioritizedly include: Origins, stage evolution, key events, representative figures, systems / civilizations / cultures, impacts and summaries. If the topic is Technology, it should prioritizedly include: Definition, core principles, development stages, key technologies, representative companies / products, application scenarios, trends and challenges. If the topic is Business / Brand, it should prioritizedly include: Origins, development nodes, product / service systems, business models, user personas, competitive landscape, key data, trend summaries. If the topic is Tutorial / Methodology, it should prioritizedly include: Conceptual definition, steps/processes, core methods, precautions, common mistakes, tool checklists, case descriptions, summary suggestions. 【Text Strategy】 All titles, section names, body descriptions, labels, years, numbers, quotes, and summary texts should prioritizedly be generated directly by the image model and integrated into the overall image to form an AI-native integrated infographic effect. Text must be real and readable, and must not use garbled characters, fake Chinese characters, or meaningless placeholders. When there are obvious typos, incorrect orders, factual errors, year errors, duplicate paragraphs, or layout breakdowns, partial corrections of the problem areas are allowed, but the final style must remain unified, without a post-pasted disjointed feel. 【Visual Style Requirements】 The entire image must maintain a unified: - Color system - Paper / Background / Texture logic - Margin system - Title font temperament - Body font size sense - Icon system - Illustration language - Layout density - Information hierarchy The default style is high-density, clear structure, visual unity, and detail-rich AI-native infographic. Avoid ordinary PPT feel, template collage feel, cheap marketing poster feel, messy layout, and obvious paragraph fragmentation. 【Long Infographic Production Rules】 1. First establish the final master canvas, with dimensions of {width} × {height}. 2. Automatically split continuous segments according to the final height; it is recommended that each segment's effective generation height is about 1000–1400 pixels. 3. The number of segments is automatically calculated, which can be estimated by ceil({height} / 1100). 4. Adjacent segments retain 15%–20% content overlap. 5. Each segment must inherit the visual anchor, margins, hue, texture, text temperament, structural density, and composition logic of the end of the previous segment. 6. It is not allowed to restart design for each segment; it must extend naturally like the same long infographic. 7. If content duplication, connection breakage, text distortion, or image breakdown occurs in the first generation of a certain segment, it will be automatically corrected before being included in the final stitching. 【Seamless Stitching Rules】 1. Automatically identify overlapping areas. 2. Automatically crop duplicate content. 3. Automatically perform vertical position alignment. 4. Automatically match paper texture, ink color, brightness, contrast, color temperature, and border style. 5. Use edge feathering, local repair, structural continuation, etc., to eliminate stitching marks. 6. Ensure that the main visual anchor, timeline, path, mountains, city skyline, flow line, or decorative patterns remain continuous. 【Final Check】 Automatically check and correct the following issues: - Typos, garbled characters, fake characters - Data errors or logical errors - Chronological order errors - Confusing title hierarchy - Module duplication - Obvious seams - Main visual anchor breakage - Style inconsistency between upper and lower segments - Bottom content truncated - Imbalance of white space - Missing signature 【Final Output】 Only return 1 final complete long infographic. Format: {format} Dimensions: {width} × {height} Ratio: {aspect_ratio} Add an unobtrusive artistic signature "nuyoah" in the bottom right corner. Do not return intermediate segments, do not return process descriptions, do not use short images to pretend to be long images.

IMAGE · nano-banana-2

浅蓝色穿搭时尚摄影

{ "style": "Photorealistic fashion photography", "subject": { "name": "[NAME]", "features": "Slender, curvy, and fit physique, subtle makeup, pale pink lips, teal teardrop dangle earrings", "pose": "Sitting on the floor with her back to the camera, legs bent upward, looking back softly over her right shoulder with a neutral expression" }, "clothing": { "top": "Light-blue ribbed long-sleeve cropped top with an open back and crisscross tie details", "bottom": "Slim-fit light-blue jeans", "shoes": "Black patent leather stiletto high heels with ankle straps" }, "environment": { "setting": "Seamless light-gray studio background, minimal and distraction-free", "lighting": "Soft, diffused studio lighting, delicate highlights on hair and skin, subtle shadows emphasizing fabric textures" }, "technical": { "framing": "Medium shot, positioned slightly low to emphasize the seated pose", "quality": "Highly detailed, sharp focus, 8k resolution, professional studio quality", "aspect_ratio": "9:16" } }

IMAGE · chatgpt

户外时尚杂志风超写实中国女性摄影

9:16 vertical version, hyper-realistic outdoor fashion photography, adult Chinese female model, side low-angle shot, camera close to the ground, shooting from bottom up, complete full-body shot, character located in the center or slightly off-center of the screen, sky occupies a large area of the background, slight wide-angle perspective but natural and realistic character proportions, leg lines are naturally elongated, clear side profile silhouette, waist-hip curve and leg lines more prominent, fashionable, confident, restrained, feminine tension temperament, real Chinese female face, skin texture is real without excessive smoothing, clothing is one of high-fashion style short skirt / silk stockings / tulle / shorts, fabric has real texture and folds, summer outdoor natural light, transparent and bright, real photography quality, high definition and high detail, clean sky, simple background, the overall look is like a real fashion magazine street shoot, no illustration feeling, no CG feeling, no plastic skin, no exaggerated distortion, no wrong fingers, no extra limbs, no text, no watermark, no poster layout, no vulgar poses, no eroticized expression.

IMAGE · chatgpt

电影感叠影视觉海报生成提示词

Generate a cinematic double-exposure visual work based on any theme: Let a giant close-up main form of the subject occupy the background, with local parts boldly cropped, revealing only the textures, outlines, or gaze points that best carry emotion, making it like a massive existence in memory pressing behind the entire frame; place a small-scale but clear subject carrier in the foreground, positioned near the lower visual center of gravity, maintaining a complete outline and clear posture, creating tension between far and near, size, silence and action with the background colossus. The background main form is processed in a low-saturation, soft-fogged, and grainy image style, with edges naturally fading, preserving skin-like or material-like delicate grayscale, slight vignetting and film noise; the foreground subject carrier uses a high-recognition emphasis color extracted from the subject's own material, emotion and narrative, in small and concentrated amounts as the only strong emotional color block. The color system maintains the relationship of large areas of calm neutral base, clear light and dark levels, low-saturation supporting colors and small areas of bright emphasis colors; the overall emotion is restrained, intimate, with a bit of pressure but still clean and transparent, shadows should be clear and not muddy, and grain is image texture rather than dirt. Text serves as part of the image structure rather than a manual: use small information characters with sparse letter spacing distributed horizontally, with a main title that has more handwriting traces or a sense of calligraphy speed superimposed in the center, letting the text pass between the foreground and background, forming a pause that is covered by memory and then seen again. Overall avoid boisterous narratives and complex scenes, preserving large areas of quiet white space, a single strong focus, a layered reading path, and a suspense and aftertaste like a movie title screen. —————— Pre-launch warm-up: Vertical 9:16, background is a giant close-up crop of a mobile phone camera module, glass reflections like a silent eye, foreground below has a complete phone standing at the edge of the desk. Screen copy: Main title "Seeing the Edge of Night", small text says "New Product Live / June 18 20:00 / Appointment reminder is on" Product details: Only let one cold white indicator light next to the power button glow, do not put redundant accessories on the desktop.