提示詞庫

Nano Banana 2 提示詞

15 條實測提示詞,涵蓋商品、建築、人像、海報等領域。

價格對比

比 Nano Banana 2 官方 API 最多便宜 35%

立即試用
解析度Google 官方 APIVidney省下
1K$0.067$0.044省 35%
2K$0.101$0.073省 28%
4K$0.151$0.116省 23%

以 Creator 年繳計算。官方價格是 Google 公布的 Nano Banana 2 API 定價,2026年9月5日核對。

Exploded product view with inner mechanics revealed

揭開內部結構的產品爆炸圖

為什麼挑這條 只要填一個佔位符,就能在無接縫白背景上得到工程圖般的爆炸視角,今年被轉貼最多的產品廣告樣式。

提示詞庫
Create an exploded products with inner mechanics [product], high-end product advertising, white seamless background, exploded view with inner mechanics revealed, outer shell hovering above core, micro screws and components suspended, perfect alignment guides implied, crisp soft shadow, ultra realistic, macro product photography, 100mm lens look, f/8, 8k, 1:1

要替換的位置 [product]

模式
文字生圖
比例
1:1
解析度
1K
花費
每張 $0.044

訣竅 這張樣片的 [product] 填的是霧面黑無線耳機充電盒。

來源 提示詞出自 Amira Zairi (@azed_ai) · 經 MeiGen.ai (jau123/nanobanana-trending-prompts) 收錄

CC BY 4.0 · 由 MeiGen.ai (jau123/nanobanana-trending-prompts) 授權。提示詞原文未經改動,也不代表對 Vidney 的背書。

立即試用

不需要 API key,也不需要訂閱

Turn a messy phone snapshot into a clean listing photo

把隨手拍的手機照變成乾淨的商品圖

為什麼挑這條 去掉拿著商品的手和雜亂背景,把主體放到純白底上,還補了接地陰影。對電商賣家來說是最立即好用的一種修圖。

提示詞庫
Identify the main product in the uploaded photo (automatically removing any hands holding it or messy background details). Recreate it as a premium e-commerce product shot . Subject Isolation : Cleanly extract the product, completely removing any fingers, hands, or clutter . Background : Place the product on a pure white studio background (RGB 255, 255, 255) with a subtle, natural contact shadow at the base to ground it. Lighting : Use soft, commercial studio lighting to highlight the product's texture and material. Ensure even illumination with no harsh glare. Retouching : Automatically fix any lens distortion, improve sharpness, and color-correct to make the product look brand new and professional .
模式
以圖生圖
比例
1:1
解析度
1K
花費
每張 $0.044
輸入
A casual snapshot of a product; hands or clutter in the frame are fine as long as the subject is clear (your own photo or a royalty-free product image).

訣竅 這張樣片的前照是用本模型的文字生成圖片自製的,不是網路上找的照片,也不是實拍。

立即試用

不需要 API key,也不需要訂閱

Isometric miniature city card with live weather

帶當日天氣的等角迷你城市卡

為什麼挑這條 Google 官方示範用的提示詞。把 London 換成任何城市,模型會一次畫出地標、天氣和上方整組文字。

提示詞庫
Present a clear, 45° top-down isometric miniature 3D cartoon scene of London, featuring its most iconic landmarks and architectural elements. Use soft, refined textures with realistic PBR materials and gentle, lifelike lighting and shadows. Integrate the current weather conditions directly into the city environment to create an immersive atmospheric mood. Use a clean, minimalistic composition with a soft, solid-colored background. At the top-center, place the title "London" in large bold text, a prominent weather icon beneath it, then the date (small text) and temperature (medium text). All text must be centered with consistent spacing, and may subtly overlap the tops of the buildings.
模式
文字生圖
比例
1:1
解析度
1K
花費
每張 $0.044

訣竅 這段文字裡的城市已經填成 London,重用時把 London 換成你要的城市即可(要替換的是城市名)。

來源 提示詞出自 Google · 取自模型官方文件

立即試用

不需要 API key,也不需要訂閱

A room pulled apart in 3D, every surface still lit

拆解成立體的房間,每個面仍在發光

為什麼挑這條 室內爆炸軸測圖,原本的光源穿過拆開的縫隙繼續打光。看起來像建築系的展板,內容完全由佔位符決定。

提示詞庫
[ROOM TYPE] pulled apart in 3D, every surface and object  floating at its exact position, slightly separated.  EXPLOSION: ceiling lifted, four walls sliding outward,  floor dropped, furniture floating in place,  objects orbiting their surfaces, light sources  still emitting across the exploded void.  The ghost of the room readable in the arrangement.  Labels on architectural elements only: "[SURFACE]" / "[MATERIAL]" / "[DIMENSION]"  LIGHTING: each element still lit by its original source,  windows casting light across the gap, lamps throwing cones  into empty space.  BACKGROUND: deep dark, the room floats in void. TITLE: "[ROOM NAME]" / SUBTITLE: "[BUILDING TYPE] [PERIOD]"  Mood: architectural model that became a universe. 4K, extraordinary spatial clarity, tack sharp.

要替換的位置 [ROOM TYPE] [SURFACE] [MATERIAL] [DIMENSION] [ROOM NAME] [BUILDING TYPE] [PERIOD]

模式
文字生圖
比例
16:9
解析度
1K
花費
每張 $0.044

訣竅 這張樣片把七個佔位符填成 1930 年代的飯店閱覽室:格子天花、胡桃木護牆板、6.4 公尺跨距,標題是 The Reading Room, Grand Hotel, 1931。

來源 提示詞出自 Alexandra Aisling (@AllaAisling) · 經 MeiGen.ai (jau123/nanobanana-trending-prompts) 收錄

CC BY 4.0 · 由 MeiGen.ai (jau123/nanobanana-trending-prompts) 授權。提示詞原文未經改動,也不代表對 Vidney 的背書。

立即試用

不需要 API key,也不需要訂閱

Casual selfie into a studio business headshot

自拍照變成棚拍的商務形象照

為什麼挑這條 提示詞把鏡頭、三點打光、看得見毛孔的膚質都寫死了,所以出來像真的進棚拍過,而不是套了磨皮濾鏡。

提示詞庫
Keep the facial features of the person in the uploaded image exactly consistent . Dress them in a professional navy blue business suit with a white shirt, similar to the reference image. Background : Place the subject against a clean, solid dark gray studio photography backdrop . The background should have a subtle gradient , slightly lighter behind the subject and darker towards the edges (vignette effect). There should be no other objects. Photography Style : Shot on a Sony A7III with an 85mm f/1.4 lens , creating a flattering portrait compression. Lighting : Use a classic three-point lighting setup . The main key light should create soft, defining shadows on the face. A subtle rim light should separate the subject's shoulders and hair from the dark background. Crucial Details : Render natural skin texture with visible pores , not an airbrushed look. Add natural catchlights to the eyes . The fabric of the suit should show a subtle wool texture.Final image should be an ultra-realistic, 8k professional headshot.
模式
以圖生圖
比例
3:4
解析度
1K
花費
每張 $0.044
輸入
A front-facing selfie or half-body casual photo with clear, unobstructed facial features.

訣竅 這張樣片的前照是用本模型的文字生成圖片自製的,不是網路上找的照片,也不是實拍。

立即試用

不需要 API key,也不需要訂閱

Relight a flat portrait into golden-hour film

把平淡的人像重打成黃金時刻的底片調

為什麼挑這條 一句話重打光的經典做法。臉不變,只換時段,加上 Kodak Portra 的顆粒和窗邊的暖調。前後並排就不用解釋了。

提示詞庫
Keep the facial features of the person in the uploaded image exactly consistent . Style : A cinematic, emotional portrait shot on Kodak Portra 400 film . Setting : An urban street coffee shop window at Golden Hour (sunset) . Warm, nostalgic lighting hitting the side of the face. Atmosphere : Apply a subtle film grain and soft focus to create a dreamy, storytelling vibe. Action : The subject is looking slightly away from the camera, holding a coffee cup, with a relaxed, candid expression. Details : High quality, depth of field, bokeh background of city lights.
模式
以圖生圖
比例
3:4
解析度
1K
花費
每張 $0.044
輸入
An ordinary indoor or flatly lit portrait; dull lighting is fine, this prompt exists to fix exactly that.

訣竅 這張樣片的前照是用本模型的文字生成圖片自製的,不是網路上找的照片,也不是實拍。

立即試用

不需要 API key,也不需要訂閱

Promotional poster with three tiers of real text

三層文字都真的能讀的宣傳海報

為什麼挑這條 測的是大家真正在意的那件事:標題、優惠貼紙、頁尾小字有沒有拼對,以及有沒有落在照片的景深裡。

提示詞庫
Design a professional promotional poster for a [Coffee Shop]. Composition : A cinematic close-up of a steaming cup of cappuccino on a rustic wooden table, autumn leaves in the background (cozy atmosphere). Text Integration :
1. Main Title : 'Autumn Special' written in elegant, gold serif typography at the top.
2. Offer : 'Buy One Get One Free' clearly displayed in a modern badge or sticker style on the side.
3. Footer : 'Limited Time Only' in small, clean text at the bottom. Quality : Ensure all text is perfectly spelled, centered, and integrated into the image's depth of field.

要替換的位置 [Coffee Shop]

模式
文字生圖
比例
3:4
解析度
1K
花費
每張 $0.044

訣竅 這張樣片的 [Coffee Shop] 保留原文面的示範值,換成任何店名或品牌名都可以。

立即試用

不需要 API key,也不需要訂閱

Vintage patent filing for any invention you name

用你指定的發明做出的老專利文件

為什麼挑這條 只填一個欄位,就得到一整頁 1800 年代的專利文件:編號圖示、鋼筆註記、發黃起斑的紙張、火漆印章。也是一次密度很高的文字渲染壓力測試。

提示詞庫
A vintage patent document for {argument name="invention" default="INVENTION"}, styled after late 1800s United States Patent Office filings. The page features precise technical drawings with numbered callouts (Fig. 1, Fig. 2, Fig. 3) showing front, side, and exploded views. Handwritten annotations in fountain-pen ink describe mechanisms. The paper is aged ivory with foxing stains and soft fold creases. An official embossed seal and red wax stamp appear in the corner. A hand-signed inventor's name and date appear at the bottom. The entire image feels like a recovered archival document—authoritative, historic, and slightly mysterious.

要替換的位置 {argument name="invention" default="INVENTION"}

模式
文字生圖
比例
3:4
解析度
1K
花費
每張 $0.044

訣竅 這張樣片的發明欄填的是機械打字機。

來源 提示詞出自 Alexandra Aisling (@AllaAisling) · 經 YouMind OpenLab (YouMind-OpenLab/awesome-nano-banana-pro-prompts) 收錄

CC BY 4.0 · 由 YouMind OpenLab (YouMind-OpenLab/awesome-nano-banana-pro-prompts) 授權。提示詞原文未經改動,也不代表對 Vidney 的背書。

立即試用

不需要 API key,也不需要訂閱

Exploded vertical infographic of any dish

任何一道菜都能做的直式分解資訊圖

為什麼挑這條 一道菜的各層在白底上分開浮起,配上指示線標籤。換一個字就能做漢堡、拿鐵或一碗麵。

提示詞庫
Create a hyper-realistic exploded vertical infographic composition of [FOOD].
At the top, characteristic particles, droplets, or garnishes associated with the food, frozen mid-air.
Below it, the primary surface or topping layer with rich texture and realistic material detail.
Underneath, the core ingredient layer showing structure, depth, and natural variation.
Below that, the supporting base or interior layer, slightly separated to reveal construction.
At the bottom, a minimal plate, bowl, cup, or base appropriate to the food’s presentation.
Pure white background, soft studio lighting, subtle shadows beneath each floating element, ultra-sharp focus, DSLR macro photography, clean infographic text labels with thin pointer lines, premium editorial food photography aesthetic, 8K quality.

要替換的位置 [FOOD]

模式
文字生圖
比例
3:4
解析度
1K
花費
每張 $0.044

訣竅 這張樣片的 [FOOD] 填的是雙層起司漢堡。

來源 提示詞出自 TechieSA (@TechieBySA) · 經 MeiGen.ai (jau123/nanobanana-trending-prompts) 收錄

CC BY 4.0 · 由 MeiGen.ai (jau123/nanobanana-trending-prompts) 授權。提示詞原文未經改動,也不代表對 Vidney 的背書。

立即試用

不需要 API key,也不需要訂閱

Levitating stacked food advertising on seamless white

白底上懸浮堆疊的食品廣告

為什麼挑這條 只是把關鍵字排在一起的短文面,卻能穩定做出包裝食品廣告常見的高調懸浮食物。試錯成本低,出得好不好一眼就看得出來。

提示詞庫
[subject], premium food advertising, white seamless background, high key studio lighting, floating stacked composition, levitating pieces, [ingredient_bits] scattered, clean minimal layout, crisp soft shadow on surface, ultra realistic, macro product photography, 100mm lens look, f/8 sharp focus, subtle texture detail, editorial commercial, 8k

要替換的位置 [subject] [ingredient_bits]

模式
文字生圖
比例
1:1
解析度
1K
花費
每張 $0.044

訣竅 這張樣片的 [subject] 填藍莓鬆餅疊層,[ingredient_bits] 填藍莓與奶油捲。

來源 提示詞出自 Amira Zairi (@azed_ai) · 經 MeiGen.ai (jau123/nanobanana-trending-prompts) 收錄

CC BY 4.0 · 由 MeiGen.ai (jau123/nanobanana-trending-prompts) 授權。提示詞原文未經改動,也不代表對 Vidney 的背書。

立即試用

不需要 API key,也不需要訂閱

Tiny 3D figures bagged like a collectible drop

像收藏品一樣裝袋的迷你 3D 公仔

為什麼挑這條 四行文字就做出盲盒商品照:透明袋、緞帶、印字吊牌。看起來就是實拍,佔位符裡塞什麼題材都吃得下。

提示詞庫
A high-resolution photograph of a clear plastic drawstring bag placed on a light gray background. Inside the bag are multiple tiny 3D [subject] figures arranged neatly. The bag is tied with a soft white ribbon and has a black label tag that reads ‘[LABEL TEXT]’. Soft lighting and clean shadows emphasize the realistic textures and details

要替換的位置 [subject] [LABEL TEXT]

模式
文字生圖
比例
1:1
解析度
1K
花費
每張 $0.044

訣竅 這張樣片的 [subject] 填太空人,[LABEL TEXT] 填 MINI CREW。

立即試用

不需要 API key,也不需要訂閱

Napkin wireframe into a high-fidelity app mockup

餐巾紙上的線框稿變成高擬真 App 設計稿

為什麼挑這條 模型會把塗鴉讀成元件:方框變成漸層按鈕,線條變成文字區塊,整個畫面再放進手機外框裡。前後差距不會看錯。

提示詞庫
Transform this rough wireframe sketch into a high-fidelity UI design mockups for a mobile app. Design System : Apply a modern, clean aesthetics similar to iOS 18 or Material Design 3 . Use rounded corners, soft drop shadows, and a vibrant primary color. Components : Intelligently interpret the sketch: turn scribbles into high-quality placeholder images , convert rough rectangles into proper buttons with gradients , and turn lines into realistic text blocks . Layout : Ensure perfect padding and consistent spacing between elements. Context : Place the design inside a realistic iPhone 16 frame mockups.
模式
以圖生圖
比例
3:4
解析度
1K
花費
每張 $0.044
輸入
A hand-drawn app wireframe sketch (on paper or a whiteboard, either works) with recognizable boxes and text lines.

訣竅 這張樣片的前照是用本模型的文字生成圖片自製的,不是網路上找的照片,也不是實拍。

立即試用

不需要 API key,也不需要訂閱

Flat isometric illustration sized for a web hero

尺寸剛好給首屏主視覺用的平面等角插畫

為什麼挑這條 SaaS 到達頁常見的粉彩等角向量風格,底是白色或淡漸層,可以直接放進首屏主視覺。

提示詞庫
A flat isometric digital illustration of [describe the subject: e.g., a modern workspace, a city block, a group of app icons, a sports shop], clean lines and geometric forms, bright pastel colors, simplified perspective with 3D depth, minimal shading, white background or light gradient. Style resembles modern vector infographics, ideal for UI, app design or web visuals.

要替換的位置 [describe the subject: e.g., a modern workspace, a city block, a group of app icons, a sports shop]

模式
文字生圖
比例
16:9
解析度
1K
花費
每張 $0.044

訣竅 這張樣片的題材欄填的是有書桌、筆電、層架和盆栽的現代居家工作區。

立即試用

不需要 API key,也不需要訂閱

Clear the tourists out of your holiday photo

把旅遊照片裡的路人清掉

為什麼挑這條 刪掉背景的人群,再用石板路、長椅或草地把空缺補回來,顆粒和對焦深度都跟周圍一致。不用解釋就看得懂的示範。

提示詞庫
Remove all the tourists/people in the background behind the main subject. Intelligent Fill : Replace them with realistic background elements that logically fit the scene (e.g., extend the cobblestone pavement, empty park benches, or grass textures). Consistency : Ensure no blurry artifacts or 'smudges' remain. The filled area must have the same grain, focus depth, and lighting as the rest of the photo.
模式
以圖生圖
比例
4:3
解析度
1K
花費
每張 $0.044
輸入
A group photo at a landmark with a clear subject and passersby cluttering the background.

訣竅 這張樣片的前照是用本模型的文字生成圖片自製的,不是網路上找的照片,也不是實拍。

立即試用

不需要 API key,也不需要訂閱

Expand a cramped photo into a 16:9 wallpaper

把太緊的照片延伸成 16:9 桌布

為什麼挑這條 往左右補畫,光線、天氣和質感都跟原圖一致,被畫面邊緣切掉的東西也會接著畫完。構圖太擠的照片就這樣變成能用的橫幅或桌布。

提示詞庫
Zoom out and expand this image to a 16:9 aspect ratio (computer wallpaper size). Context Awareness : Seamlessly extend the scenery on both left and right sides. Match the original lighting, weather, and texture perfectly. Logical Completion : If there are cut-off objects (like a shoulder, a tree branch, or a building edge) on the borders, complete them naturally based on logical inference. Do not distort the original center image.
模式
以圖生圖
比例
16:9
解析度
1K
花費
每張 $0.044
輸入
A tightly framed photo (landscape or portrait both work) with scenery at the edges that can plausibly be extended.

訣竅 這張樣片的前照是用本模型的文字生成圖片自製的,不是網路上找的照片,也不是實拍。

立即試用

不需要 API key,也不需要訂閱

這個模型做不到的

跑過但沒有收進來的提示詞,每一條都用同樣的方式失敗了兩次。

把 2D 平面圖變成室內設計提案板

提示詞庫
Based on the uploaded 2D floor plan, generate a professional interior design presentation board in a single image. Layout : The final image should be a collage with one large main image at the top, and several smaller images below it. Content of Each Panel :
1. Main Image (Top) : A wide-angle perspective view of the main living area , showing the connection between the living room and dining area.
2. Small Image (Bottom Left) : A view of the Master Bedroom , focusing on the bed and window.
3. Small Image (Bottom Middle) : A view of the Home Office / Study room .
4. Small Image (Bottom Right) : A 3D top-down floor plan view showing the furniture layout. Overall Style : Apply a consistent Modern Minimalist style with warm oak wood flooring and off-white walls across ALL images. Quality : Photorealistic rendering, soft natural lighting.

怎麼失敗的 兩次都在 3D 平面圖的房間標籤上出錯。第一次把 BATH 拼成 BATN,尺寸變成 5.7in a 4.0m;第二次把 KITCHEN 拼成 KITCHCN,主臥尺寸變成一串亂碼字元,右側還多出一張被裁切的重複平面圖。版面數量也沒有做到提示詞要求的一大三小。

一張自拍生成九種表情的 3D 貼圖組

提示詞庫
Using the uploaded selfie as the ONLY and exclusive face reference, generate a Pixar-style 3D character sticker set. The character must be clearly recognizable as the person from the selfie. Create a neat 3×3 grid collage (nine stickers total). Final aspect ratio is strictly 4:5. Each cell shows a distinct pose and facial expression. Style and tone: Pixar-style 3D animation with a meme-oriented feel. Exaggerated, slightly absurd emotions, playful overacting. 

Highly readable expressions at small sizes, with expressive eyes, eyebrows, and mouth. Subtle hand gestures where appropriate. Background and output: Fully transparent background (PNG). No backdrop, no background shadows. No text, captions, logos, or UI elements. Stickers must be ready for direct use in messengers. Consistency: Same hairstyle, clothing, colors, and character proportions across all nine stickers. Only facial expression and pose change.

Clean, polished 3D sticker look. Emotions (one per sticker, 3×3 grid): Surprised, Annoyed, Confused, Frustrated, Thoughtful, Sarcastic, Worried, Bored, Curious.

怎麼失敗的 兩次都沒有做出透明背景。模型把代表透明的灰白棋盤格當成圖案直接畫進畫面,輸出本身也沒有 Alpha 通道。第二次還違反了提示詞明確寫下的不要文字、不要說明字,在九格下方各加了一行情緒名稱。

來源 提示詞出自 Oogie (@oggii_0) · 經 MeiGen.ai (jau123/nanobanana-trending-prompts) 收錄

CC BY 4.0 · 由 MeiGen.ai (jau123/nanobanana-trending-prompts) 授權。提示詞原文未經改動,也不代表對 Vidney 的背書。

這個需求該換誰

  • 畫面裡密集的小字,我們會交給 Nano Banana Pro。同一個要求換到那邊還有機會。 Nano Banana Pro 模型頁

常見問題

這些提示詞可以免費使用嗎

可以。複製後在任何地方跑都行。每張卡片都標了原作者與公開時的授權條款,你可以自己核對。我們自己寫的那幾條,無條件開放再利用。

要有帳號才能跑嗎

只複製提示詞完全不需要帳號。在 Vidney 上生成才需要,而新帳號都附贈免費點數,頭幾次不花錢。

方括號裡的字是什麼

那是要你替換的位置。換成自己的題材,其餘句子原樣保留,因為撐住結果的正是周圍那些措辭。

一張圖要多少錢

每張卡片都顯示旁邊那組設定要花的點數,不是手寫上去的,而是從即時價目表讀出來的。改解析度,花費也跟著變。

為什麼還列出失敗的提示詞

知道天花板跟知道地板一樣有用。同樣方式失敗兩次的,我們不會悄悄刪掉,而是寫清楚怎麼失敗的,留在那一節裡。

整頁 15 條全跑一遍,大約 $0.65