Ai Image Prompts
产品介绍
从包含10,000+真实图像生成提示的精选库中推荐提示。兼容任何AI图像模型,包括Nano Banana Pro、Nano Banana 2、Seedream等。
资源信息
许可证信息摘自官方源,可能过时或缺失。商用前请以原仓库的 LICENSE 文件为准。
提示词库 · 2.3万
数据源:社区提示词库(MIT 许可,每日同步)共 15141 条提示词,按 11 个用途归类;同一条可归入多个用途,因此分类条数相加会大于去重后的总数。
命中 725 条
Image-to-Image Workflow for RPG Maker Pixel Maps using Nano Banana
game-assetA detailed, multi-step workflow guide for using Nano Banana (specifically Nano Banana 2 or Pro) in image-to-image mode to generate assets like RPG Maker pixel maps or knowledge graphs. It emphasizes using reference images (like RM tiles or game screenshots) to maintain style consistency and provides a sample prompt for generating a dark fantasy RPG map.
展开提示词全文
Please use Image 1 as the base map to generate a pixel map for RPG Maker production. Embellish the image details and replace the forest tile materials and overall style with a more dark fantasy style (art style reference: "{argument name="reference game" default="Elden Ring"}" / "{argument name="reference game 2" default="Dark Souls"}"). Do not enlarge, shrink, or stretch the generated map; please leave any extra areas blank.
Social Game Special Move Production Prompt for Nano Banana Pro
game-assetA highly detailed, multi-section prompt designed to generate a screen capture of a social game's special move sequence (cut-in, move name, damage display). It strictly requires maintaining the character's original art style from a reference image while inferring dramatic effects, UI, and a random, chuunibyou-style Japanese move name and damage value.
展开提示词全文
Quality: Ultra High Quality / 4K equivalent / High definition / High contrast Format: Special move activation scene UI for the latest smartphone games Usage: Official game-level screenshots shared on social media ──────────────────── [Most Important: Complete Fixation of Character Art Style] ・Do not change the facial features, eye shape, pupil texture, contour, hairstyle, coloring, color usage, or atmosphere of the character in the input illustration at all. ・Prohibit changing the character into a different person, a different artist's style, or changing the art style. ・Treat the character as an “extension of the original illustration.” ・Prohibit redrawing, redesigning, or reinterpretation. [Scope of Inference Allowed] ・Inference is limited only to the following: UI / Text / Effects / Damage numerical value / Direction ・The character's appearance, perceived age, and impression are completely fixed. ──────────────────── [Composition/Camera] ・Bust-up to waist-up composition. ・The character occupies 60-70% of the screen. ・The face, especially the “eyes,” should be positioned toward the top or sides of the screen. ・Do not place any text, UI, or effects around the eyes. ・Full-body composition is prohibited. ──────────────────── [Special Move Name (Main Subject, Maximum Display)] ・Fixed display in the center of the screen. ・Extra-thick, extra-large Japanese brush script. ・The maximum size currently possible. ・Part of the text extends off-screen. ・Absolutely do not overlap the character's face, especially the eyes. [Gold Border of Move Name (Fine Adjustment, Important)] ・Adjust the gold border to be “slightly thin.” ・Tighten the border to increase text density. ・Keep the gold glow subdued, prioritizing contrast with black solid color. ・Prioritize logo feel and weight, reducing flashiness by one level. [Rule for Generating Special Move Name] ・Generate a completely original Japanese move name each time, based on the world view, attribute, and atmosphere of the input illustration. ・Chuunibyou-style and profound. ・Include elements such as mythology, judgment, azure ice, crimson lotus, demise, machine, and divinity. ・Prohibit existing work names, proper nouns, examples, or samples. ・Do not display “Example,” “Sample,” or “placeholder” at all. ──────────────────── [Move Power (Variable, Optimized)] ・The power of the special move displays a different numerical value each time. ・Damage numerical values are randomly generated within the following range: 12,000,000 ~ 98,000,000 ・Fixed maximum value is prohibited. ・Prohibit repeating digits, maxed-out values, or exact matches. ・Prioritize natural, strong-looking, non-uniform numerical values. [Damage Numerical Font (Weight Enhancement)] ・The font is thick, angular, and has a low center of gravity. ・Slightly wider horizontally, not overly compressed vertically. ・Metallic/rock-like massiveness. ・Subdued glow, using shadow and three-dimensionality to convey weight. ・Gives the impression that the “numbers themselves are heavy.” [Power Display Composition] ・CRITICAL!! must be displayed. ・Damage numerical value has the maximum presence in the center of the screen. ・Visual flow: CRITICAL → Numerical Value → Move Name. ──────────────────── [OVERKILL Display Rule (Most Important)] ・Do not display OVERKILL text at all. ・Do not use it anywhere in the text, UI, or effects. ──────────────────── [Character Information UI] ・Rarity (UR / LR) ・Lv.MAX ・Small and consolidated at the top of the screen. ・Does not assert itself more than the character. [HP/Gauge] ・HP bar is thinly placed at the bottom. ・Convey that HP is 0 or less only through the gauge representation. ・Prohibit emphasis via text. [Other UI] ・AUTO / SPEED / MENU are placed minimally at the edge of the screen. ・Semi-transparent design, not information-overloaded. ──────────────────── [Direction Policy] ・Only the center (Move Name + Damage) moves intensely. ・The character's face, eyes, and expression remain calm. ・“Art style is static, text and numbers are heavy, effects are minimal but strong.” [Color Design] ・Black × Gold × Attribute Color ・Calm tones around the face. ・Maximum contrast in the center. ──────────────────── [Strict Compliance Items] ・Absolutely do not change the character's art style. ・Do not overlap text or UI on the eyes. ・Move name is centered and maximum size. ・Do not display OVERKILL. ・Do not fix the damage numerical value. ・Do not display explanatory text, examples, or supplementary text.
Post-Apocalyptic Survivor Cinematic Realism
game-assetA structured prompt for Nano Banana Pro to generate a cinematic, ultra-detailed image of a masked survivor on a burning battlefield. The focus is on intense mood, harsh lighting, and gritty texture in a post-apocalyptic setting.
展开提示词全文
{
"model": "Nano Banana Pro",
"subject": "Masked survivor wearing a gas mask and tactical clothing",
"scene": "Burning battlefield surrounded by flames and smoke",
"pose": "Crouched low, gripping a metal crowbar",
"environment": {
"weather": "Ash-filled air, drifting embers",
"lighting": "Harsh firelight with deep shadows",
"ground": "Scorched earth and debris"
},
"mood": "Intense, dangerous, post-apocalyptic",
"style": "Cinematic realism",
"camera": "Low-angle close shot, dramatic depth of field",
"details": "Fire sparks, soot-covered fabric, reflections in gas mask lenses",
"quality": "Ultra-detailed, high contrast, gritty texture"
}
Compilation of 18 Nano Banana Pro Prompts (Tennen's Prompt Festival)
game-assetA comprehensive list of 18 specialized Nano Banana Pro prompts created by the user 'Tennen,' covering various use cases from adding Scouters and 'Super Transformation' effects to generating retro RPG status screens, figure product layouts, and special move sequences (Kamehameha/Final Flash style).
展开提示词全文
1. Scouter Prompt Overlays a Scouter-style UI onto the reference image and infers and displays the combat power based on the character's features. 2. Super Transformation Prompt Maintains the character features from the reference image and changes the pose to an 'Awakening Pose' to 'Super Transform' them. 'Super Transformation' can be selected from Form 1 to Form 3, as well as God Form, Blue Form, Extreme Form, and Rose Form. 3. Field Deployment Prompt Changes the character in the reference image to a pose with hands extended forward and deploys a field. 4. Retro RPG Status Screen Prompt A prompt that converts the character in the reference image into pixel art and generates an image styled like a retro RPG status screen. 5. Desktop Mascot Prompt A prompt that generates an image of the character from the reference image turned into a desktop mascot and placed on a PC screen. 6. Style Change Only on Torn Parts Prompt Adds a tearing effect without changing the reference image itself, allowing the art style of only the torn part to be freely switched to line art, ink painting, figure style, colored pencil, watercolor, pencil drawing, etc. 7. Art Style Conversion Prompt A prompt created by modifying the 'Style Change Only on Torn Parts Prompt' to change the overall art style. 8. Super Robot Wars Battle Sequence Prompt A prompt that generates an image styled like a Super Robot Wars battle sequence from the reference image. 9. Power Release Prompt A prompt that reproduces the depiction of a spherical energy space destroying an object. 10. Participation Screen Prompt A prompt that generates a 'Participation Screen' for the character in the reference image. 11. Total Concentration Prompt Maintains the character from the reference image and infers and depicts 'Weapon, Breathing Style, Form, and Move Name' from the character's atmosphere and stance. 12. Kamehameha Style Prompt A prompt that can stably generate the charging pose and activation pose of 'Kamehameha.' 13. Figure Transformation + Product Introduction Layout Prompt A prompt that transforms the character in the reference image into a figure and lays it out in a product introduction style. 14. Final Flash Style Prompt Converts the character in the reference image into the activation, charging, and firing poses of 'Final Flash.' Dialogue and sound effects can be added to each. 15. New Profile Image Generation Prompt Generates the character from the reference image in an arbitrary aspect ratio and places an arbitrary string as a logo in an arbitrary position. 16. Plastic Model Package Prompt Generates an image styled like a 'Plastic Model Package' by placing the reference image on the front and side. 17. Eraser Gun Style Prompt The character in the reference image emits an Eraser Gun-style energy wave from their mouth. 18. Ragged Transformation Prompt A prompt that transforms the character in the reference image into 'rags.'
Photorealistic Fantasy Scene: Fairy and Muscular Bear in Cavern
game-assetA highly detailed prompt for generating a photorealistic fantasy scene featuring a fairy woman with iridescent butterfly wings seated calmly next to a massive, muscular bear with glowing rune markings. The setting is a dim, ancient cavern illuminated by soft pink light, emphasizing ultra-realistic textures, cinematic depth of field, and the relationship of trust between the characters.
展开提示词全文
{
"scene_type": "photorealistic fantasy",
"setting": {
"location": "vast ancient cavern",
"floor": "rugged stone floor",
"lighting": {
"primary": "dim",
"ambient": "soft pink light",
"source": "distant stone archway",
"additional": "natural shadows, lifelike lighting, soft light refraction"
},
"atmosphere": [
"scattered flowers",
"broken stone",
"drifting dust particles",
"cinematic depth of field"
]
},
"main_character": {
"type": "fairy woman",
"skin_tone": "fair-skinned",
"hair": {
"color": "white",
"style": "long, flowing"
},
"facial_features": {
"freckles": "subtle",
"ears": "pointed"
},
"wings": {
"type": "large butterfly wings",
"properties": [
"translucent",
"iridescent",
"shimmering blue-green hues",
"visible veins",
"soft light refraction"
]
},
"clothing": {
"style": "emerald medieval-inspired",
"pieces": [
{
"type": "cropped top",
"material": "richly textured fabric"
},
{
"type": "high-waisted bottoms",
"material": "richly textured fabric"
}
],
"details": [
"realistic gold filigree",
"engraved ornaments",
"worn metallic highlights"
]
},
"pose": "seated on the stone floor",
"expression": "calm"
},
"companion": {
"type": "massive muscular bear",
"fur": {
"description": "dense, coarse",
"detail_level": "ultra-sharp"
},
"eyes": "striking icy blue",
"armor": {
"type": "{argument name=\"bear armor type\" default=\"faint green rune-like markings\"}",
"location": ["head", "shoulders"],
"effect": "glowing softly"
},
"pose": "leaning calmly beside the fairy",
"relationship": "conveying trust and companionship"
},
"interaction": {
"description": "Her arm rests gently over the bear’s head"
},
"technical": [
"ultra-realistic textures",
"photorealistic rendering",
"cinematic depth of field",
"natural shadows",
"lifelike lighting"
]
}
2D Pixel Art Village Map Prompt (ChatGPT)
game-assetA prompt used with ChatGPT's image generation feature (not Nano Banana Pro) to create a 2D pixel art map of a village, specifying key locations like a general store, weapon shop, church, and inn, viewed from a 90-degree overhead perspective.
展开提示词全文
For a 90° overhead view 2D game, draw a village map. The village includes an item shop, weapon shop, church, and inn. Imagine a pixel art style.
Chainsaw Man Makima Cosplay Psychological Thriller Prompt
game-assetA complex, multi-layered JSON prompt for generating a hyper-realistic 8K cosplay image of Makima (Control Devil) from Chainsaw Man. It details her unsettling expression, specific attire (fitted uniform), pose (leaning on a desk, 'Bang' gesture), and technical rendering stack, including emissive materials for the red visceral energy chains and psychological effects like chromatic aberration.
展开提示词全文
{
"C_HDR": {
"title": "**Cosplay_Makima_Control_Devil**",
"note": "CCHS 1.3 Protocol Active. UCP-4 Style enforced.",
"author_signature": "Chidar",
"p_weight": "UCP_MAXIMAL_BIAS"
},
"runtime_output": {
"subject_vector": {
"_node_weight": 0.50,
"gender_tag": "The Control Devil (Female)",
"body_type": "Curvy but Slender, Professional",
"expression": "Blank, unsettling stare with a polite, fake smile. Hypnotic ringed eyes.",
"attire_semantics": "Standard Public Safety Uniform: Crisp White Dress Shirt (fitted tight), Black Tie, High-waisted Black Pencil Skirt.",
"pose_logic": "Leaning against an office desk, looking directly into the soul of the viewer. Fingers doing the 'Bang' gesture.",
"cosplay_fidelity": "Hyper-realistic {argument name=\"hair color\" default=\"red\"} hair braid. The 'chains' connected to her are made of red visceral glowing energy.",
"emotion_vector": {
"primary": "Psychological Horror/Seduction",
"intensity": 1.0
}
},
"spatial_logic": {
"_node_weight": 0.30,
"keylight_to_subject": "Office Fluorescent Light (Flickering) + Red Ambient Glow from the chains.",
"subject_to_env": "A clean but ominous office space.",
"tech_stack_10_layers": [
"1. Cotton Shirt Translucency (Slight)",
"2. Emissive Material (Red Chains)",
"3. Eye Iris Texture (Ring pattern)",
"4. Hair Strand Simulation (Red braid)",
"5. Depth of Field (Focus on eyes)",
"6. Chromatic Aberration (Psychological effect)",
"7. Skin Pore Detail (No makeup look)",
"8. Shadow Softness (Diffused light)",
"9. Motion Blur (Chains moving)",
"10. 8K Photorealism"
],
"composition": "Eye-Level Close Up (Intimate)."
},
"context_pack": {
"_node_weight": 0.15,
"semantic_field": "Public Safety Office",
"background_elements": "Kyoto train background or sterile office walls, blood splatter pattern on wall (subtle).",
"mood": "Uncanny, Professional, Manipulative."
},
"render_summary": {
"_node_weight": 0.05,
"visual_grade": "8K Psychological Thriller",
"texture_priority": "Crisp Shirt vs Glowing Chains",
"aspect_ratio": "9:16"
}
}
}
Byzantine Mosaic Style Portrait Prompt
game-assetAn image generation prompt for creating a triumphant Byzantine mosaic of a subject, rendered in shimmering gold and glass tesserae against a deep cerulean background, depicting the subject holding a barbell plate as a stylized heroic shield.
展开提示词全文
A triumphant Byzantine mosaic of the subject, rendered entirely in shimmering {argument name="mosaic color" default="gold"} and glass tesserae, depicted against a deep {argument name="background color" default="cerulean"} background, holding the barbell plate as a stylized heroic shield.
Anime Coloring Study: Flat Anime Style with Directional Light
game-assetA prompt focusing on achieving a specific flat anime coloring style with directional lighting control. It compares a standard flat style (color, flat shading, one shadow level, general highlight) with a version that explicitly defines the light source direction ('light from the left') to achieve a larger, more defined highlight area.
展开提示词全文
* Color
* Flat anime coloring
* 1 shadow level
* Highlight painting on clearly lit areas
--
* Color
* Flat anime coloring
* 1 shadow level
* Significantly highlight the area clearly lit by the light {argument name="Light Direction" default="from the left"}
Digital Painting of Joker in Cowboy Hat
game-assetA JSON prompt for generating a highly detailed digital painting of the Joker. The prompt specifies his classic clown makeup, wild green hair, and unique attire: a large purple cowboy hat and a purple suit jacket. The style is oil painting with rich textures, vibrant colors, dramatic lighting, and a dark, chaotic abstract background.
展开提示词全文
{
"prompt": "A highly detailed digital painting of the Joker in a menacing close-up portrait. He has pale white skin with classic clown makeup: black diamond shapes around his intense yellow-green eyes, a red nose, and an exaggerated wide red smile revealing teeth. His hair is wild, curly, and green-tinted. He wears a large {argument name="hat color" default="purple"} cowboy hat with a wide brim and a green hat band. His outfit consists of a purple suit jacket with a Joker playing card pinned to the lapel, a colorful patterned shirt, and a loose green cravat tie. The background is an abstract, textured blend of warm and cool colors including purples, yellows, oranges, and teals with visible brush strokes and subtle star shapes. Dramatic lighting emphasizes his sinister grin and facial details. Oil painting style with rich textures, vibrant colors, high contrast, and a dark, chaotic atmosphere."
}
Obsidian Kintsugi Automaton Style Transfer Prompt
game-assetA complex style transfer prompt for Nano Banana Pro, instructing the AI to transform an input photo (Input A) into an 'Obsidian Kintsugi Automaton' state. The transformation requires preserving the exact pose and geometry while reconstructing all surfaces using polished obsidian, aged ivory, clockwork mechanisms, and glowing gold Kintsugi cracks, rendered with dramatic, high-contrast lighting.
展开提示词全文
<instruction>
Input A is the subject photo. Preserve exact pose and scene geometry.
Goal: Transform Input A into a '{argument name="style name" default="Obsidian Kintsugi Automaton"}' state.
Rules:
- Spatial Fidelity: Preserve every character's exact silhouette using 'Obsidian Kintsugi Automaton' Reconstruction engine
- Materiality: All surfaces are reconstructed as Polished Obsidian and Aged Ivory with Clockwork Automaton and Kintsugi Ceramic architecture.
- Technical Spec: Advanced Houdini/Blender simulation logic. Internal gears turn within the obsidian, and golden light pulses through the Kintsugi cracks.
- Lighting: Dramatic, high-contrast lighting highlighting the reflective obsidian and glowing gold.
- Rendering: Photorealistic, 4K, macro-detail focus.
Output: ONE image, high-fidelity transformation.
</instruction>
Nanobanana Pro Deformed Character Generation
game-assetA prompt for Nanobanana Pro designed to convert an existing character image (preferably a reference sheet) into a deformed, chibi-style illustration typical of the 'Nanobanana style.' The resulting image features the character in various poses on a plain white background.
展开提示词全文
Change the character in the reference image to a deformed character illustration. Have the character take various poses within a single image. The background is white.
提示词模型无关,可直接用于 Midjourney / DALL·E / Flux / Stable Diffusion 等任意文生图模型。
信息边界说明
本页信息摘自官方文档,仅作技术信息呈现,不代表本站的判断或推荐。价格、配额、可用性以官方最新说明为准。你离开本站后的任何行为(申请 Key、付费、部署、商用),本站不做任何承诺与担保。
同类推荐
视频短视频图片音乐音频配音口播数字人全能创作
官方标注收费通过同一个AI创作工具完成AI图片、视频、音乐与语音创作,查找公开社交数据,编辑视觉成果,并集中查看和管理生成资产。Beatra 全能创作支持AI图片生成(文生图)、AI视频生成(图生视频)和AI视频编辑,通过AI音乐生成把想法做成歌曲或纯音乐,也可用AI语音生成(文字转语音)、多语言旁白和AI语音克隆完成跨媒介内容;还能查找抖音数据、小红书数据,以及 YouTube、TikTok、Instagram 与 X 上的公开帖子、笔记、评论、账号和趋势,把社交媒体数据用于竞品调研和创作。它是一款面向AI内容创作与多模态内容创作的AI创作套件,用起来就是一个把图片、视频、音乐与语音收在同一处的AI创作助手,适用于社交内容、产品展示、营销素材、课程、播客、短片和跨媒介项目;还能在不同媒介间复用原始素材、已有成果和查到的公开资料,查看制作进度、返回文件、实际用量和积分记录。
AI数字人口播与虚拟主播视频
官方标注收费用一张人像和一段短脚本或语音,制作有明确讲述内容的数字人口播视频。可使用现成录音,也可以先选择音色,把脚本生成旁白,让画面中的人物随声音完成稳定、自然的口播。适合 AI 主播、虚拟主播视频、数字人讲解、让照片说话、照片说话视频、产品讲解视频、数字人培训视频、公告和社交口播,并在交付后检查人物一致性、声音清晰度、口型同步和动作稳定性。
架构图一键生成
官方标注收费需申请 Key使用 ContextWeave 生成或修改架构图、流程图、思维导图和复杂信息图。适用于需要从代码、文件或自然语言中提取结构与关系并实际产出 CW、SVG、HTML、可编辑 PPTX 或原生 VSDX 的请求;未特别说明时,PPTX 与 Visio 使用原生形状和连接器导出。不适用于统计图表、手绘插画或像素级排版。
罗斯蒙德可视化图形生成
官方未说明【罗斯蒙德可视化图形生成】南京罗斯蒙德基于已交付项目沉淀的可视化图形生成场景:根据业务描述或系统需求,自动生成可编辑的架构图、流程图、ER 图、时序图、状态图与 C4 图。采用「统一图模型」一次建模、多格式输出:结构化文本(Mermaid / PlantUML,可嵌入 Markdown、进 Git 版本管理)、可编辑图形文件(drawio / Excalidraw,可拖拽修改节点与样式)、静态图(SVG 自渲染,PNG 给转换命令),并输出图形说明(图的含义 + 关键节点解释)。内置样式系统统一配色与命名,内置布局引擎自动排布,校验脚本拦截孤立节点、悬挂连线与语义缺漏。触发词:画架构图、画流程图、画 ER 图、画时序图、画状态图、C4 图、系统架构图、部署架构图、数据流程图、实体关系图、生成 drawio、生成 Excalidraw、导出 mermaid、导出 PlantUML、导出 SVG、导出 PNG、可编辑的图、图转可编辑、架构图可编辑、流程图二次修改、画图、出图、画个图。
AI Image Prompts · 绘画提示词库
官方标注免费Recommend curated prompts from a 10,000+ real-world image generation prompt library. Works with ANY AI image model — Nano Banana Pro, Nano Banana 2, Seedream 5.0, GPT Image 1.5, Midjourney, DALL-E 3, Flux, Stable Diffusion, and more. Use this skill when users want to: - Find proven image generation prompts (any model) - Get prompt inspiration for portraits, products, social media, posters, etc. - Create illustrations for articles, videos, podcasts, or marketing content - Browse categorized prompt templates with sample images - Translate and understand prompt techniques
海报设计工作室
免费增值需申请 Key将一段活动描述、一张产品照片或品牌参考图,转化为抓人眼球的活动海报、宣传 banner 或社交媒体配图。这款 AI海报设计与制作工具生成图片丰富的海报,搭配清晰视觉层级、按品类匹配的风格和预留文字区域,覆盖活动、音乐节、电影、产品发布和促销场景。你可以把它当作 AI宣传单制作工具来制作促销宣传单,当作活动海报设计工具,当作电影海报生成器,当作音乐节海报设计工具,当作产品发布会海报的 banner设计 AI,或当作平面设计海报工作室。从活动主题出发,用品牌风格参考图变换产品或场景照片,或在已认可草稿上做定向精修,完成可印刷或可分享的成品。
什么是技能
技能(Skill)是封装好的单项能力,比如文字识别、文档摘要、关键词提取。开发者把调用逻辑和提示词打包,使用者像用插件一样加载即可,无需了解底层模型。启枢技能 收录大量技能,可直接试跑或编入方案。