提示词库
Nano Banana 2.1 提示词，每条附生成的图。
每条提示词都原样摘自模型厂商的指南或创作者的帖子，旁边是它生成的成品。点开卡片看全文。
Google 在 2026 年 10 月 6 日发布 Nano Banana 2.1，这是它的图像生成和编辑模型，API 里叫 gemini-nano-banana-2.1，定位是 Nano Banana Pro 的高效版。它接受文字、图片、视频和 PDF 输入，输出 1K、2K 或 4K 的图，新增 1:4、4:1、1:8 和 8:1 画幅，最多接受 14 张参考图，其中物体最多 10 张、人物最多 4 个。API 价格：1K 图每张 0.0336 美元，2K 0.0504 美元，4K 0.113 美元，走 Batch API 减半。下面的提示词都是创作者在 X 上分享的，每条附成图并链到原帖。
来源：Google on X · Gemini API: Nano Banana 2.1 · Image generation guide · Pricing
高赞的 Nano Banana 2.1 提示词怎么写：81 条帖子实测 →
Nano Banana Pro 和 2.1 的提示词比一比 →
GPT Image 2.6 和 Nano Banana 2.1 的提示词比一比 →
更新于 2026-11-03 · 共 90 条，最新的在前
其他模型页：GPT Image 2.6 · Seedance 2.5 · 可灵 4.0 · MiniMax H3 · 万相 3.0
橄榄绿西装男子，皮肤质感自然
门铃摄像头拍到的糖果怪物
十个穿雨衣的同款双胞胎比手指数字
古庙台阶上女子的反向视角
穿精灵服的浣熊偷吃饼干被抓现行
粉色卧室中的镜子自拍，女子手臂带有纹身
黄色串珠挂颈上衣与湿发妆造人像
蜘蛛侠十种画风标签拼图
霓虹黄缎面迷你裙女子坐于天鹅绒扶手椅
复古厨房里手持提拉米苏的金发模特
深蓝雾气影棚中的深色连帽衫电影感肖像
镜面走廊中穿白色缎面吊带裙的女子
Base 周报按板块排列的标志方块
Maggi 2 Minute Magic 日出街头面条广告牌
桑达尔·皮查伊从5岁到现在的人生阶段时间线
四位漫画家的玩具手办，作品从包装盒中跃出
窗边阳光下身穿白T恤的手机自拍
逐项复刻一张照片的 JSON 提示词
情侣衫的线头抽成一幅线稿
日本新年家宴上的动画风家庭冲突
穿Polo衫与复古球鞋的动漫街头风情侣
复古混凝土廊道中的酒红色时装人像
圆形洞口中的混凝土楼梯黑白人像
黄金时刻屋顶男装人像
黑色套装男子在电梯里自拍
黑色大衣男子与埃菲尔铁塔街拍
湖畔露台上身着礼服啜饮茶的男子
少女与小猫的三格可爱人像拼贴
半脸真实照片半脸波普艺术肖像
手机质感的四宫格美妆人像
四位科学家姓氏的实验性字标网格
四本讽刺时事的立体书与黏土机器
石阶上扎头发的夜间闪光抓拍
硬侧光黑白时尚编辑人像
带动态模糊虚影的编辑风人像海报
金色落日积云前的骑马牛仔
海岸金色时刻的男性特写肖像
砂岩墙前的静奢风男装人像
极简奢华广告风侧脸墨镜人像，发丝飞扬
温室里龟背竹叶后的斑驳光影人像
黄昏海滩上身穿粉色丝绸长裙的赤脚女子
照片中仅头发、碗和筷子变成动漫插画
卧室梳妆台前直视镜头的金发博主
四个概念词化作符号造型的字体插画
颜料管挤出立体厚涂画作的四宫格
夜间街头旋转人像，慢快门动态模糊
青绿色复古咖啡馆的鱼眼镜面自拍
没有匹配的条目，换个短一点的词试试。
值得照着改的 Nano Banana 2.1 提示词
本页每个分类里点赞最多的一条，全文贴出，并说明它为什么好用。复制一条，换掉主体就能用。
窗边阳光下身穿白T恤的手机自拍
用结构化JSON把面部、头发、服装、光线和相机拆成独立字段，每个细节都能单独控制。写明24到28毫米手机镜头、过曝高光，再配上很长的避免清单，画面保持随手拍的真实感，不会变成影棚质感。
{
"prompt_type": "detailed_image_reconstruction",
"aspect_ratio": "16:9 horizontal portrait",
"overall_goal": "Recreate the visual appearance, framing, lighting, styling, and casual smartphone-photo feel of the reference image as closely as possible while keeping the result photorealistic and natural.",
"subject": {
"type": "young adult woman",
"age_range": "early to mid 20s",
"pose": "standing indoors, upper body turned slightly away from the camera while her face turns toward the lens",
"body_position": "relaxed casual posture with shoulders slightly angled",
"expression": "soft neutral expression with a subtle pout, calm and confident",
"gaze": "looking directly toward the camera",
"styling": "minimal everyday styling, clean and natural"
},
"face": {
"shape": "soft oval face with a gently tapered jawline",
"forehead": "medium-height forehead partly framed by center-parted hair",
"eyebrows": "natural medium-thick dark brown eyebrows with a soft arch",
"eyes": {
"shape": "almond-shaped",
"color": "light brown to hazel",
"appearance": "bright eyes with visible catchlights",
"lashes": "long defined eyelashes with subtle mascara"
},
"nose": "small straight nose with a softly rounded tip",
"lips": {
"shape": "full lips with a defined cupid's bow",
"color": "natural rosy pink",
"finish": "soft satin finish, not glossy"
},
"skin": {
"tone": "light warm-neutral skin tone",
"texture": "smooth natural skin with subtle pores and realistic tonal variation",
"cheeks": "soft natural pink flush",
"finish": "fresh luminous skin without heavy retouching"
},
"makeup": "very light natural makeup, softly defined lashes, subtle blush, natural pink lips, no dramatic contour"
},
"hair": {
"color": "rich medium-to-dark chocolate brown",
"length": "very long, extending well below the chest",
"parting": "clean center part",
"texture": "mostly straight with very soft natural bends through the lower lengths",
"density": "thick and healthy-looking",
"finish": "smooth and glossy but still realistic",
"front_sections": "two thin face-framing strands falling naturally along both sides of the face",
"details": "visible individual strands, subtle flyaways near the crown, realistic highlights from sunlight, natural uneven strand separation",
"movement": "hair hanging naturally over both shoulders and down the front of the shirt"
},
"clothing": {
"top": "oversized plain white crew-neck T-shirt",
"fit": "loose casual fit",
"fabric": "soft slightly wrinkled cotton",
"logos": "none",
"accessories": "no visible jewelry or prominent accessories"
},
"composition": {
"shot_type": "close upper-body portrait",
"camera_angle": "slightly above chest level, close to eye level",
"camera_distance": "approximately arm's-length smartphone distance",
"subject_position": "face slightly right of center with long hair occupying much of the lower frame",
"crop": "top of head fully visible, image extending down to around lower torso",
"orientation": "vertical",
"perspective": "natural wide smartphone-camera perspective with mild close-range distortion"
},
"lighting": {
"primary_source": "strong warm natural sunlight entering from the side/front through a nearby window",
"direction": "sunlight striking the subject from camera-left/front",
"face_effect": "bright warm illumination across the forehead, cheeks, nose, and hair",
"hair_effect": "distinct brown highlights and soft reflective streaks along individual strands",
"shadow_quality": "strong but soft-edged natural indoor shadows",
"exposure": "face slightly brightly exposed while the room behind remains much darker",
"color_temperature": "warm daylight"
},
"environment": {
"location": "casual indoor residential room",
"background": "dark gray walls with a white ceiling cornice",
"left_background": "open doorway or darker hallway area",
"right_background": "dark wall and partially visible household objects",
"lower_right": "part of a black tripod or light stand visible",
"overall_feel": "ordinary lived-in room rather than a studio",
"depth": "background moderately out of focus but still recognizable"
},
"camera": {
"capture_device": "modern smartphone camera",
"lens_equivalent": "approximately 24-28mm full-frame equivalent",
"focus": "sharp focus on face and front hair",
"depth_of_field": "mostly natural smartphone depth of field with gentle background softness",
"dynamic_range": "realistic phone-camera HDR with slightly clipped highlights in direct sun",
"image_character": "casual social-media selfie/photo, not professionally staged"
},
"photorealism_details": [
"real individual hair strands",
"natural skin pores and micro-texture",
"subtle peach fuzz",
"realistic eyelash separation",
"tiny variations in lip texture",
"slight fabric creases in the T-shirt",
"small flyaway hairs around the crown",
"natural sunlight falloff across the face",
"slightly imperfect casual framing",
"realistic indoor shadow noise"
],
"style": {
"visual_style": "highly photorealistic casual smartphone portrait",
"mood": "soft, youthful, relaxed, sunlit",
"processing": "light social-media style processing with warm skin tones and slightly enhanced clarity",
"avoid": [
"cinematic movie lighting",
"studio portrait lighting",
"overly smooth plastic skin",
"CGI appearance",
"anime or illustration",
"heavy beauty filter",
"extreme facial symmetry",
"over-sharpened hair",
"artificial jet-black hair",
"dramatic makeup",
"excessive background blur",
"professional studio backdrop",
"text",
"watermarks",
"logos"
]
},
"negative_prompt": "cartoon, anime, illustration, CGI, plastic skin, waxy face, over-retouched skin, excessive beauty filter, unnatural facial symmetry, distorted eyes, crossed eyes, malformed lips, warped face, duplicate hair strands, fake wig appearance, overly black hair, metallic hair shine, stiff hair, messy anatomy, distorted shoulders, extra limbs, extra fingers, studio backdrop, cinematic lighting, dramatic color grading, excessive bokeh, oversaturated colors, text, logo, watermark"
} 质感几何风全身时装插画
服装逐件描述，再用一小串禁止项排除写实和 3D。
Create a stylish, full-body editorial fashion illustration of a young woman in a modern casual outfit, using a minimalist, textured geometric art style. She wears an oversized forest-green hoodie, wide-leg blue denim jeans with folded cuffs, and chunky brown sneakers. Her dark hair is tied in a neat low bun. She carries a bouquet of vibrant pink roses wrapped in white paper, a takeaway coffee cup, and a black shoulder bag. Show her walking naturally in a relaxed pose, viewed from a three-quarter rear angle, with soft, elegant body proportions and simplified facial details.
Use a warm ivory background with plenty of negative space, soft ground shadows, subtle paper grain, visible brush textures, and angular polygonal shading across the clothing and shoes. The color palette features deep forest green, denim blue, warm brown, cream, and vivid pink. Soft diffused lighting, clean composition, contemporary lifestyle magazine aesthetic, hand-painted digital illustration, refined edges, artistic fabric textures, vertical 3:4 composition, high detail. No photorealism, no 3D rendering, no extra objects, no text or watermark. 一半照片一半水墨的海报
要求严格五五分，再分别规定两半如何处理同一主体。
Using the uploaded reference image, create a single high-resolution vertical 2:3 editorial art poster.
Divide the canvas into two perfectly equal horizontal sections, each occupying exactly 50% of the total height. Separate them with a clean, perfectly straight horizontal dividing line.
1. TOP HALF — ORIGINAL REFERENCE PHOTOGRAPH
Use the uploaded reference photograph directly in the upper 50% of the poster.
Preserve its original photographic quality, sharpness, natural lighting, and color relationships.
Allow only subtle, intelligent cropping around the outer edges when necessary, without compromising the main subject.
If required, naturally extend a small portion of the surrounding environment to fit the composition.
Do not redraw, illustrate, stretch, blur, or modify the subject's appearance, structure, or spatial relationships.
Do not introduce black borders, white margins, or empty edges.
2. BOTTOM HALF — MINIMALIST EDITORIAL ILLUSTRATION
Use a warm, near-white ivory paper background with an extremely subtle paper texture.
Analyze the main subject and visual story of the uploaded reference image.
Extract only the primary subject, preserving one or two essential identifying elements.
Remove or heavily simplify all secondary details, including background architecture, crowds, vehicles, fences, signs, trees, and other unnecessary elements.
Transform the primary subject into a sophisticated black-ink editorial illustration.
ILLUSTRATION STYLE:
Use natural, clean pen strokes with varied line weight.
Preserve an authentic hand-drawn appearance with expressive but controlled linework.
Avoid overly precise engraving aesthetics, excessive internal details, and dense cross-hatching.
Place a large, nearly pure-black rectangular ink block behind the illustrated subject.
The rectangular block should have subtly imperfect, handmade print-style edges.
Keep the black background clean, with only fine paper grain and minimal noise.
Avoid obvious hatching, scratches, architectural lines, or complex textures.
ILLUSTRATION SCALE AND COMPOSITION:
The illustrated subject must occupy approximately 72–77% of the canvas width and 58–64% of the bottom section's height.
Maintain a clear visual focal point with generous negative space surrounding the illustration.
3. COLOR RULES
The bottom illustration must follow a strictly black-and-white color palette by default.
Preserve only one localized accent color if the original reference image contains a clearly distinctive, concentrated color element with meaningful narrative importance.
Ordinary blue skies, green vegetation, stone surfaces, skin tones, and environmental colors must not be treated as accent colors.
Any preserved accent color must remain strictly localized to the relevant subject or area.
Never allow accent colors to spread across the entire illustration.
4. TYPOGRAPHY AND EDITORIAL LAYOUT
Automatically generate concise, meaningful English typography based on the content of the uploaded reference photograph.
The typography must resemble handwritten editorial annotations from an independent art publication.
TOP-LEFT CORNER:
Generate a short handwritten title consisting of 2–5 words, arranged across two lines.
Directly beneath the title, include three short English action verbs relevant to the photograph.
TOP-RIGHT CORNER:
Display only two lines of geographic coordinates:
Latitude
Longitude
Do not include any additional text in this area.
BOTTOM-LEFT CORNER:
Display the location name followed by the word "DIARY."
Directly beneath it, add one short descriptive English sentence inspired by the photograph.
TYPOGRAPHY STYLE:
Use a rough, upright, slightly irregular handwritten typeface.
Maintain subtle imperfections and a natural editorial annotation aesthetic.
Avoid elegant cursive, calligraphy, overly decorative lettering, and excessively polished typography.
5. FINAL ART DIRECTION
The finished poster must combine:
Authentic, high-quality photography.
Minimalist black-and-white editorial illustration.
Generous negative space.
Independent art publication aesthetics.
Quiet, restrained, sophisticated visual storytelling.
Natural handwritten typography.
A balanced relationship between documentary realism and expressive hand-drawn artwork.
The overall result should resemble a carefully designed, collectible editorial art print.
6. FINAL OUTPUT REQUIREMENTS
ASPECT RATIO:
2:3 vertical.
RESOLUTION:
High-resolution, sharp, print-quality output.
COMPOSITION:
Exactly 50% original reference photography on top.
Exactly 50% editorial illustration on the bottom.
One perfectly straight horizontal dividing line.
One complete, standalone poster.
RESTRICTIONS:
No collages.
No multi-poster layouts.
No combined images.
No unnecessary decorative elements.
No modifications to the original subject's identity or appearance.
No excessive illustration details.
No changes to the original photograph's natural color relationships.
FINAL GOAL:
Create a sophisticated, visually balanced editorial art poster that faithfully preserves the uploaded reference photograph while transforming its primary subject into an expressive, minimalist black-ink illustration. 逐项复刻一张照片的 JSON 提示词
按 JSON 字段写（构图、光线、色彩、负面词）。原帖用同一段 JSON 对比了 Nano Banana 2.1 和 GPT Image 2.5。
{
"prompt_type": "detailed_image_reconstruction",
"aspect_ratio": "4:5 vertical portrait",
"overall_goal": "Recreate the visual appearance, framing, lighting, styling, and casual smartphone-photo feel of the reference image as closely as possible while keeping the result photorealistic and natural.",
"subject": {
"type": "young adult woman",
"age_range": "early to mid 20s",
"pose": "standing indoors, upper body turned slightly away from the camera while her face turns toward the lens",
"body_position": "relaxed casual posture with shoulders slightly angled",
"expression": "soft neutral expression with a subtle pout, calm and confident",
"gaze": "looking directly toward the camera",
"styling": "minimal everyday styling, clean and natural"
},
"face": {
"shape": "soft oval face with a gently tapered jawline",
"forehead": "medium-height forehead partly framed by center-parted hair",
"eyebrows": "natural medium-thick dark brown eyebrows with a soft arch",
"eyes": {
"shape": "almond-shaped",
"color": "light brown to hazel",
"appearance": "bright eyes with visible catchlights",
"lashes": "long defined eyelashes with subtle mascara"
},
"nose": "small straight nose with a softly rounded tip",
"lips": {
"shape": "full lips with a defined cupid's bow",
"color": "natural rosy pink",
"finish": "soft satin finish, not glossy"
},
"skin": {
"tone": "light warm-neutral skin tone",
"texture": "smooth natural skin with subtle pores and realistic tonal variation",
"cheeks": "soft natural pink flush",
"finish": "fresh luminous skin without heavy retouching"
},
"makeup": "very light natural makeup, softly defined lashes, subtle blush, natural pink lips, no dramatic contour"
},
"hair": {
"color": "rich medium-to-dark chocolate brown",
"length": "very long, extending well below the chest",
"parting": "clean center part",
"texture": "mostly straight with very soft natural bends through the lower lengths",
"density": "thick and healthy-looking",
"finish": "smooth and glossy but still realistic",
"front_sections": "two thin face-framing strands falling naturally along both sides of the face",
"details": "visible individual strands, subtle flyaways near the crown, realistic highlights from sunlight, natural uneven strand separation",
"movement": "hair hanging naturally over both shoulders and down the front of the shirt"
},
"clothing": {
"top": "oversized plain white crew-neck T-shirt",
"fit": "loose casual fit",
"fabric": "soft slightly wrinkled cotton",
"logos": "none",
"accessories": "no visible jewelry or prominent accessories"
},
"composition": {
"shot_type": "close upper-body portrait",
"camera_angle": "slightly above chest level, close to eye level",
"camera_distance": "approximately arm's-length smartphone distance",
"subject_position": "face slightly right of center with long hair occupying much of the lower frame",
"crop": "top of head fully visible, image extending down to around lower torso",
"orientation": "vertical",
"perspective": "natural wide smartphone-camera perspective with mild close-range distortion"
},
"lighting": {
"primary_source": "strong warm natural sunlight entering from the side/front through a nearby window",
"direction": "sunlight striking the subject from camera-left/front",
"face_effect": "bright warm illumination across the forehead, cheeks, nose, and hair",
"hair_effect": "distinct brown highlights and soft reflective streaks along individual strands",
"shadow_quality": "strong but soft-edged natural indoor shadows",
"exposure": "face slightly brightly exposed while the room behind remains much darker",
"color_temperature": "warm daylight"
},
"environment": {
"location": "casual indoor residential room",
"background": "dark gray walls with a white ceiling cornice",
"left_background": "open doorway or darker hallway area",
"right_background": "dark wall and partially visible household objects",
"lower_right": "part of a black tripod or light stand visible",
"overall_feel": "ordinary lived-in room rather than a studio",
"depth": "background moderately out of focus but still recognizable"
},
"camera": {
"capture_device": "modern smartphone camera",
"lens_equivalent": "approximately 24-28mm full-frame equivalent",
"focus": "sharp focus on face and front hair",
"depth_of_field": "mostly natural smartphone depth of field with gentle background softness",
"dynamic_range": "realistic phone-camera HDR with slightly clipped highlights in direct sun",
"image_character": "casual social-media selfie/photo, not professionally staged"
},
"photorealism_details": [
"real individual hair strands",
"natural skin pores and micro-texture",
"subtle peach fuzz",
"realistic eyelash separation",
"tiny variations in lip texture",
"slight fabric creases in the T-shirt",
"small flyaway hairs around the crown",
"natural sunlight falloff across the face",
"slightly imperfect casual framing",
"realistic indoor shadow noise"
],
"style": {
"visual_style": "highly photorealistic casual smartphone portrait",
"mood": "soft, youthful, relaxed, sunlit",
"processing": "light social-media style processing with warm skin tones and slightly enhanced clarity",
"avoid": [
"cinematic movie lighting",
"studio portrait lighting",
"overly smooth plastic skin",
"CGI appearance",
"anime or illustration",
"heavy beauty filter",
"extreme facial symmetry",
"over-sharpened hair",
"artificial jet-black hair",
"dramatic makeup",
"excessive background blur",
"professional studio backdrop",
"text",
"watermarks",
"logos"
]
},
"negative_prompt": "cartoon, anime, illustration, CGI, plastic skin, waxy face, over-retouched skin, excessive beauty filter, unnatural facial symmetry, distorted eyes, crossed eyes, malformed lips, warped face, duplicate hair strands, fake wig appearance, overly black hair, metallic hair shine, stiff hair, messy anatomy, distorted shoulders, extra limbs, extra fingers, studio backdrop, cinematic lighting, dramatic color grading, excessive bokeh, oversaturated colors, text, logo, watermark"
} 蜘蛛侠十种画风标签拼图
只给出一个主体和风格数量，不指定具体画风，由模型自行选定十种风格并排成带标签的网格。要求同一张图中呈现同一角色，各格的人物形象保持一致，便于直接比较画风。
🚨 New Nano Banana 2.1 just got new checkpoint on arena < high chance this is new pro version i will test and keep you guys updated> 上面照片下面素描的编辑海报模板
填 `[SUBJECT]` 和 `[SETTING]`。禁止项针对分割海报的常见毛病：粗边框、上下脸不一致。
Create a premium editorial poster featuring a two-part visual composition. The upper half shows a photorealistic photograph of [SUBJECT] in [SETTING], captured in warm natural sunlight with realistic skin texture, authentic expressions, soft highlights, and cinematic depth of field. The lower half transforms the exact same scene into a beautiful hand-painted digital illustration, preserving the subject’s pose, hairstyle, clothing, colors, and key visual details.
Place the illustrated scene inside a clean rectangular frame on a warm ivory, subtly textured paper background. Use a charming contemporary illustration style with delicate brush textures, simplified organic shapes, pastel colors, soft outlines, and a nostalgic summer-magazine aesthetic. Add tasteful decorative elements that complement the scene.
Below the illustration, include a centered minimalist editorial title reading “[MAIN TITLE]” in thin, widely spaced uppercase typography, followed by a smaller subtitle reading “[TAGLINE]” in elegant letter-spaced text. Add a short, thin pastel-colored line underneath.
Maintain balanced spacing, refined art direction, a cohesive color palette, and a sophisticated print-design finish. Vertical 3:4 composition, high resolution, natural lighting, polished editorial layout, aesthetically pleasing, realistic-to-illustration transformation.
Avoid clutter, excessive text, distorted anatomy, inconsistent facial features, harsh colors, heavy borders, and unnecessary graphic elements. 这个模型的提示词怎么写
用完整的句子描述画面
Google 的指南建议把画面写得越细越好，写得越具体，越能控制结果。它给照片的模板是：某种景别的写实照片，拍某个主体，在某个场景里，再补上光线和机位。创作者写得很长：这页 90 条提示词的中位长度是 1,749 个字符，约 250 个英文词，其中 50 条写了镜头、景别或景深。 来源：Google: Photorealistic scenes · AgentCraft: Nano Banana 2.1 and Seedance 2.5 prompt data
写清图片拿来做什么
Google 的最佳实践里有两条：交代用途，让模型知道这张图拿来做什么；复杂的画面拆成按顺序的几步来写。同样的主体，做商品图、海报还是头像，好结果的标准都不一样。 来源：Google: Best practices
改图时，点名要改的部分，再说其余不动
不用画遮罩：用文字描述要改的地方，再加一句图里其余部分完全保持原样。涉及人脸或标志时，把它详细描述一遍，并写明其特征完全不变。创作者也这么写：这页 32 条改图提示词里有 24 条写了保持、保留或不要改动，从文字生成的 58 条里只有 11 条。 来源：Google: Inpainting with semantic masking · AgentCraft: Nano Banana 2.1 and Seedance 2.5 prompt data
参考图编号，并说明每张的用途
Nano Banana 2.1 最多接受 14 张参考图，人物最多保持 4 个的相似度，物体最多保持 10 个的还原度。Google 给多图的模板是：把所给图片里的元素组合成一张新图，再说清哪个元素来自哪张。这页只有 2 条提示词给上传的图编了号，在这一点上写清楚，就比大多数人强。 来源：Google: Use up to 14 images · AgentCraft: Nano Banana 2.1 and Seedance 2.5 prompt data
写画面里应该有什么
Google 把这叫作语义上的反向提示：直接描述你要的画面，少列要排除的东西。这页 90 条里仍有 51 条在某处写了不要、避免或没有；结果里总出现你禁掉的东西时，就把那一句改写成正面描述。 来源：Google: Best practices · AgentCraft: Nano Banana 2.1 and Seedance 2.5 prompt data
要出现的字，写清字体和版式
要在图上出现文字，Google 建议写清原文、字体风格和整体设计；它在已知限制里还写到，最好先让模型把文字定下来，再要求生成含这段文字的图。这页有 25 条提示词把文字放进了引号。 来源：Google: Accurate text in images · Google: Limitations
先定尺寸和画幅，再小步修改
默认尺寸是 1K，也可以选 2K 和 4K，K 要大写。这页有 49 条提示词在正文里写了画幅比例。出了第一张之后，Google 建议在后续对话里小步修改，不必整条重写。 来源：Google: Aspect ratios and image size · Google: Best practices
关于这个模型的常见问题
Nano Banana 2.1 是什么？
Google 在 2026 年 10 月 6 日发布的图像生成和编辑模型。Google 说它全面超过之前的图像模型，提升最大的是视觉设计、按区域改图和主体一致性。API 名称是 gemini-nano-banana-2.1。 来源：Google on X
Nano Banana 2.1 的 API 怎么收费？
按张算：1K 0.0336 美元，2K 0.0504 美元，4K 0.113 美元，折合每百万图片 token 30 美元。输入每百万 token 1.5 美元。Batch API 全部减半，没有免费额度。作个对比，Nano Banana Pro 的 1K 或 2K 图约 0.134 美元一张，4K 0.24 美元。 来源：Gemini API pricing
Nano Banana Pro 和 Nano Banana 2.1 选哪个？
大多数情况选 2.1：每张图不到 Pro 的一半价，物体参考图能放 10 张，Pro 是 6 张，还多了 1:4 到 8:1 的画幅。要保持更多人物一致（Pro 5 个，2.1 是 4 个）、要用最多 3 张风格参考图，或者做文字多的专业物料时选 Nano Banana Pro（Gemini 3 Pro Image），Google 的指南在这种情况下推荐它。 来源：Google: Use up to 14 images
最多能用几张参考图？
一次最多 14 张。Google 说一次流程里最多能保持 4 个人物的相似度和 10 个物体的还原度。 来源：Google: Use up to 14 images
支持哪些尺寸和画幅？
默认 1K，另有 2K 和 4K。在常规画幅之外，新增了 1:4、4:1、1:8 和 8:1，用于长条竖图和全景图。Gemini 3.1 Flash Image 的 512px 尺寸在 2.1 上不能用。 来源：Google: Aspect ratios and image size
能不画遮罩，只改照片的一部分吗？
可以。Google 把这叫作语义遮罩：用文字描述要改的部分，并告诉模型图里其余部分完全保持原样。 来源：Google: Inpainting with semantic masking
提示词用哪种语言效果最好？
Google 列出了效果最好的语言，其中包括英文、日文和简体中文。这页 90 条提示词全是英文写的。 来源：Google: Limitations
Nano Banana 2.1 生成的图有水印吗？
有。Google 说所有生成的图都带 SynthID 水印。 来源：Google: Limitations
Nano Banana 2.1 的提示词该写多长？
画面需要多长就写多长。我们把 X 上 81 条 Nano Banana 2.1 提示词和点赞数对上：随机各挑一条高赞和低赞的，高赞那条更长的概率只有 49%。高赞的中位长度 1,503 个字符，低赞的 1,857 个；一条只有 59 个字符的报纸拼贴也进了前四分之一。 来源：AgentCraft: Nano Banana 2.1 prompts that get liked
高赞的 Nano Banana 2.1 提示词有什么共同点？
通过检验的只有两项：photorealistic 或 hyper-realistic 这个词在前一半里占 49%，后一半只有 25%；写成 JSON 的 6 条全在前一半，其中 4 条是人像。画幅、保持不变的说明、镜头术语、反向要求和带标题的分段，两边没有差别。 来源：AgentCraft: Nano Banana 2.1 prompts that get liked
大家用 Nano Banana 2.1 做什么？
大多是人像。这页 90 条提示词里，46 条是人像，10 条插画，8 条照片转画风，8 条写实场景，5 条海报。32 条从上传的照片开始改，这类改图提示词更长，中位数 2,066 个字符，从零写的是 1,720 个。 来源：AgentCraft: Nano Banana 2.1 and Seedance 2.5 prompt data
改照片时怎么让人脸不变？
给人脸单独写一句。从上传照片改人像的提示词里，Nano Banana 2.1 有 81% 写了锁定人脸或身份，Nano Banana Pro 是 62%。一个例子是在开头单列一段：IDENTITY LOCK: Preserve the exact facial identity of the reference person。 来源：AgentCraft: Nano Banana Pro vs 2.1 prompts
Nano Banana Pro 的提示词能直接用在 2.1 上吗？
可以先照搬。Google 的指南建议新项目都用 2.1。然后补上 Pro 提示词常漏的两样：画幅比例，2.1 的提示词有 53% 写了，Pro 只有 23%；改图时再加一句锁定人脸。 来源：AgentCraft: Nano Banana Pro vs 2.1 prompts
Nano Banana 2.1 能用 Google 搜索、读视频或 PDF 吗？
能。它接受文字、图片、视频和 PDF 输入，支持用 Google 搜索做依据，还有思考程度设置：minimal、medium 或 high，默认 medium。 来源：Gemini API: Nano Banana 2.1