智東西AI前瞻(公眾號:zhidxcomAI)
作者 | 江宇
編輯 | 漠影

智(zhi)東西(xi)9月(yue)5日(ri)報道,Google最新一代(dai)圖(tu)像生成模型Nano Banana(Gemini 2.5 Flash Image)已(yi)于近日(ri)上線,現(xian)已(yi)支持文生圖(tu)、圖(tu)文編(bian)輯、多(duo)圖(tu)合(he)成、迭代(dai)精修(xiu)以(yi)及文字渲染等能力。自此,Nano Banana在海內外社交媒體上迅速刷屏(ping):從X到小紅書(shu),用戶(hu)紛(fen)紛(fen)曬出改圖(tu)、拼圖(tu)、造(zao)場景的(de)案例(li)。

前日,Google AI Studio還總結了(le)六條提示(shi)詞(ci)技巧(qiao),其(qi)核心要義就(jiu)在于:用(yong)完整(zheng)的(de)場景敘事,而(er)不(bu)是零散的(de)關鍵詞(ci)。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲谷歌官方案例圖

比如,如果(guo)想要(yao)生成一張寫(xie)實感很強的(de)照片,可以把自(zi)己當成攝影(ying)師來(lai)寫(xie)提示詞,重點放在(zai)氛圍、鏡頭和(he)細節上。要(yao)做貼紙或插畫(hua)風(feng)格(ge),就別忘了強調線(xian)條的(de)勾勒、顏色的(de)搭配,以及純白背景(jing)的(de)干凈效(xiao)果(guo)。如果(guo)希望文字能精準嵌入畫(hua)面,那(nei)就要(yao)講清楚字體、排版層(ceng)次(ci)和(he)整(zheng)體設計風(feng)格(ge)。

要做(zuo)產(chan)品棚拍(pai)或商業廣告圖(tu),最關(guan)鍵的(de)就是打光方(fang)式、拍(pai)攝角度(du)和材(cai)質質感(gan);而極簡(jian)留白(bai)的(de)風格,則需要控(kong)制(zhi)畫(hua)(hua)面的(de)留白(bai)空間、主體的(de)位(wei)置和光線的(de)柔(rou)和度(du)。至于分鏡或漫畫(hua)(hua)格,更看重的(de)是角色動作和環境的(de)呼應,以及光影營造出來的(de)氛圍。

不過,看完這(zhe)些“官方(fang)技巧”,你可能仍覺得意猶未盡(jin)。畢竟在(zai)社交(jiao)平臺(tai)上,已經有人玩出了更(geng)多意想不到(dao)的花(hua)樣。今(jin)天,智東西就(jiu)帶來一份Nano Banana的20種玩法合集(ji),從人物合影(ying)到(dao)卡牌設(she)計、從商品植入(ru)到(dao)3D場(chang)景,一并整(zheng)理(li)。

一、人物篇(共7類)

人物篇共有7類玩法,包含生(sheng)成(cheng)雙人合(he)影、貓貓自(zi)拍、極速變裝、Cosplay、改姿(zi)勢(shi)、老(lao)照片修(xiu)復和表(biao)情(qing)包貼紙。

1、雙人合影

Prompt: A photorealistic portrait of the two people from the reference image, standing side by side as friends, one taller and one shorter, smiling warmly. Their facial features and details should remain consistent with the reference photo. They are outdoors in a park during spring blossom season, with blooming flowers in the background. The lighting is soft and natural, creating a cheerful atmosphere. Captured with a 50mm lens, highlighting their joyful expressions.

指令:寫實風(feng)格的(de)人像畫(hua)面(mian),來自(zi)參考圖片中的(de)兩(liang)位人物,并肩站(zhan)立成(cheng)朋(peng)友(you)姿態,一高一矮,面(mian)帶笑容。面(mian)部細節需與參考照片保持一致。他們身處春日公園的(de)花海中,背景(jing)有盛開的(de)花樹。光(guang)線柔和自(zi)然(ran),營造(zao)愉快氛圍。使用50mm鏡頭(tou)拍攝,突出他們的(de)喜悅神情。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參考圖(tu)(reference image)

Nano Banana能夠直接把扎克伯格和薩姆(mu)·阿爾特曼拉到(dao)同一張畫(hua)面里,兩人并(bing)肩,被(bei)“塑造”成了春日踏青的好友(you)。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

不僅,Nano Banana把“公園”和(he)“花海”等指令細節復刻得(de)相當到位(wei),還(huan)自動給兩人加了個搭肩摟腰的姿(zi)勢(shi),看起來氛圍格外“融洽”。細節上,扎(zha)克伯格的發型(xing)和(he)參考圖幾乎一模一樣,阿(a)爾(er)特(te)曼的發型(xing)稍(shao)有(you)跑偏(pian),但臉部特(te)征(zheng)還(huan)是非常精準(zhun)。

2、貓貓自拍

Prompt: A photorealistic portrait originally showing the cat with Batman, but now replace Batman with Elon Musk. Keep all other elements exactly the same as the original setting (including the cat, background, lighting, and atmosphere). Elon Musk must have the same facial features and expression as in his reference image.

指令:寫(xie)實風格的人像(xiang)合影(ying),原(yuan)本設定為小貓與蝙(bian)蝠俠的合影(ying),現在(zai)改為埃(ai)隆(long)·馬(ma)斯(si)克(ke)與小貓合影(ying)。其他元素保持完(wan)全一致(包括小貓、背景、光線和氛圍(wei))。埃(ai)隆(long)·馬(ma)斯(si)克(ke)需(xu)保持參(can)考圖中的面部(bu)特征和表情。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參考圖(reference image)

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參(can)考圖(reference image)

這張“小貓(mao)自拍”的原圖在海外已(yi)經被(bei)玩出了(le)無數版本,有蝙蝠(fu)俠,也有超人,角色的一致性都保持得很不錯。

智東西(xi)這次讓小(xiao)貓換(huan)個“搭檔”,和馬(ma)斯克來了一(yi)張(zhang)合影。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

在幾次抽(chou)卡的結(jie)果里,有一(yi)張保留了馬斯(si)克參考圖中的“搞怪(guai)”表情,效果意外有點喜感。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

3、極速變裝

Prompt: A photorealistic full-body portrait of Sam Altman, with his outfit swapped to the trench coat from the reference image. He is standing confidently in a runway/neutral studio backdrop, with the same styling and pose as the reference model. The scene is illuminated by soft, diffused spotlight, creating a modern, stylish atmosphere. Captured with a 50mm full-body portrait lens, emphasizing the fabric texture and flowing silhouette of the trench coat. The composition should be vertical, fashion photography format. Ensure the clothing and styling are consistent with the reference runway outfit. Ultra-realistic, with high clarity. Maintain a natural head-to-shoulder ratio, consistent with realistic human proportions. Ensure the head, face, and body remain well-proportioned and harmonious.

指(zhi)令(ling):寫(xie)實風(feng)格的(de)全身人像(xiang),薩(sa)姆·阿(a)爾特曼換上(shang)參(can)考(kao)(kao)圖中(zhong)的(de)風(feng)衣。他以(yi)與(yu)參(can)考(kao)(kao)模特相同的(de)姿態與(yu)造型站(zhan)立在秀場/中(zhong)性攝影棚背景中(zhong)。場景采用柔和的(de)聚光燈(deng)照(zhao)明,營造現代、時尚(shang)的(de)氛圍。畫面以(yi)50mm全身人像(xiang)鏡頭拍攝,突出(chu)風(feng)衣的(de)布料質(zhi)感與(yu)流暢廓(kuo)形。構圖為豎版時尚(shang)攝影格式。需確(que)保服(fu)(fu)裝與(yu)造型與(yu)參(can)考(kao)(kao)圖的(de)秀場服(fu)(fu)裝保持(chi)一(yi)致。成像(xiang)效果超(chao)寫(xie)實,清晰(xi)度高。保持(chi)自然的(de)頭肩(jian)比例,符合(he)真實人體的(de)比例感。保證頭部、面部與(yu)身體協調統一(yi)。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

這一(yi)條(tiao)提(ti)示(shi)詞是按照Google官方(fang)的(de)指令tips專門優化過的(de)。起初在沒有強調(diao)“頭、臉、身體(ti)比例統一(yi)”時,生成的(de)人像多少會有些(xie)別(bie)扭。加上這一(yi)層(ceng)約束后,身體(ti)比例就自然(ran)了許多。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

4、Cosplay

Prompt: A photorealistic full-body portrait of Sam Altman cosplaying as Captain America, wearing a detailed blue suit with a white star on the chest, red-and-white stripes, and holding the iconic circular shield. He is standing confidently in front of a dramatic battlefield backdrop, with stormy clouds and rays of sunlight breaking through. The lighting is cinematic and natural, highlighting the realistic textures of the fabric, metal shield, and his facial features. The overall mood is heroic and inspiring.

指令(ling):寫實(shi)風格的(de)全身(shen)人(ren)像,薩姆·阿爾(er)特曼Cos成(cheng)美(mei)國隊長,身(shen)穿細(xi)節豐富的(de)藍(lan)色制服(fu),胸前(qian)有白(bai)色星星與紅白(bai)條紋(wen),手(shou)持標(biao)志性(xing)的(de)圓形(xing)盾(dun)牌(pai)。他自(zi)信地站立在一片(pian)戰場背景前(qian),天空(kong)布滿陰云,陽光(guang)(guang)從云層間穿透。畫面光(guang)(guang)效具(ju)有電影感且自(zi)然,突出(chu)服(fu)裝面料、金屬(shu)盾(dun)牌(pai)以及面部特征的(de)真實(shi)質感。整(zheng)體氛(fen)圍英勇(yong)而(er)鼓(gu)舞人(ren)心。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

Prompt:A photorealistic full-body portrait of Elon Musk cosplaying as Batman, wearing a detailed black armored Batsuit with a flowing cape and the iconic bat symbol on his chest. He stands in front of a dark, rainy Gotham City rooftop backdrop, with neon lights reflecting off wet surfaces. The lighting is dramatic and moody, highlighting the textures of the armor, the cape’s folds, and his facial features. The overall atmosphere is intense, heroic, and cinematic.

指令:寫實(shi)風(feng)(feng)格的(de)全身人像,埃(ai)隆·馬斯克Cos成蝙蝠俠(xia),身穿(chuan)細(xi)節豐富的(de)黑色裝甲戰衣,披(pi)著飄逸(yi)的(de)披(pi)風(feng)(feng),胸(xiong)口有標志性的(de)蝙蝠徽(hui)章。他站(zhan)立在暗雨夜(ye)中(zhong)的(de)哥譚市屋頂,霓虹燈倒映在濕漉(lu)漉(lu)的(de)地面上。突出(chu)了戰衣的(de)質感、披(pi)風(feng)(feng)的(de)褶皺以及面部特征。整體(ti)氛圍緊張、英勇,極具(ju)電影感。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

對于(yu)這種Cosplay場景,Nano Banana幾(ji)乎是(shi)“信手拈來”。說(shuo)是(shi)AI生成的圖,但效(xiao)果能和幾(ji)年(nian)前的好萊塢大片媲美了。

5、改姿勢

Prompt:A stylized full-body illustration of the cartoon pig character from the reference image, performing the pose from the action reference image. The pig’s appearance, colors, and smooth cartoon style must remain exactly the same as in the original reference. Generate three camera perspectives of the same pose:

Front view–full-body, centered composition.

45-degree angle view–showing depth and dimension.

Low-angle shot–emphasizing power and dynamism.

The background, lighting, and atmosphere should remain consistent with the original pig reference image. Only the pose is changed, matching the action reference exactly.

指令:風格(ge)化全身插畫(hua),讓(rang)參考(kao)圖(tu)中(zhong)的(de)(de)(de)卡通小豬角(jiao)(jiao)色(se)擺(bai)出動作參考(kao)圖(tu)中(zhong)的(de)(de)(de)姿(zi)勢。小豬的(de)(de)(de)外觀、顏色(se)與(yu)(yu)平滑的(de)(de)(de)卡通風格(ge)需與(yu)(yu)原參考(kao)圖(tu)保(bao)持完(wan)全一(yi)致。生成該姿(zi)勢的(de)(de)(de)三個不同視(shi)(shi)角(jiao)(jiao):正(zheng)面視(shi)(shi)角(jiao)(jiao)——全身、居中(zhong)構圖(tu);45度斜角(jiao)(jiao)視(shi)(shi)角(jiao)(jiao)——展現(xian)空間感和(he)立體感;仰角(jiao)(jiao)視(shi)(shi)角(jiao)(jiao)——強調動態效果。背景、光線和(he)氛圍需與(yu)(yu)小豬參考(kao)圖(tu)保(bao)持一(yi)致。唯一(yi)改變(bian)的(de)(de)(de)是姿(zi)勢,應與(yu)(yu)動作參考(kao)圖(tu)完(wan)全匹(pi)配(pei)。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參(can)考圖(reference image)

智(zhi)東西給(gei)Nano Banana上(shang)了(le)點難度。動(dong)作參考圖的(de)人體結構相對準確,智(zhi)東西直接讓(rang)動(dong)畫(hua)電(dian)影中的(de)角(jiao)色浪(lang)浪(lang)山小豬(zhu)妖去模仿。結果生成的(de)畫(hua)面還挺讓(rang)人驚喜:小豬(zhu)妖清澈的(de)眼神和整(zheng)體形象(xiang)都(dou)(dou)復刻得很穩。三個角(jiao)度的(de)生成沒有任何問(wen)題,甚(shen)至連它標(biao)志性的(de)豬(zhu)鬃毛都(dou)(dou)保留了(le)下來,整(zheng)幅(fu)畫(hua)也(ye)依(yi)舊(jiu)保持(chi)了(le)上(shang)美廠的(de)原始畫(hua)風。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

6、老照片修復

Prompt: Colorize old black-and-white photos with realistic, historically accurate tones. After natural colorization, retain mild film grain and a soft vignette to preserve the vintage feel. Emulate Kodak/Fuji color negative film aesthetics — warm highlights, stable mid-tones, and restrained shadows. Use moderate-to-low saturation, avoiding neon or trendy filters, so the final image feels authentic, as if it could have been shot in color at that time.

指令:還原并增強(qiang)一張老舊(jiu)的黑白照片(pian)(pian),修復劃痕、折(zhe)痕和(he)褪色細節。提升畫面的清晰(xi)度與銳度,同時保留(liu)復古質感。對(dui)人物膚色、衣物和(he)背景元素進行細致而真實的上色處理,讓照片(pian)(pian)煥(huan)發新生,同時維持(chi)原始影像(xiang)的懷舊(jiu)氛圍(wei)。

這一回,智(zhi)東西把(ba)Nano Banana攻克的目標放在了英國(guo)作家蕭伯納的黑白肖像上。在修復時,為整張(zhang)照(zhao)片(pian)做了細致的上色。膚色自然、衣物的質感都(dou)被處理(li)得(de)比較貼(tie)近真實(shi)。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

7、表情包貼紙

Prompt:A set of cartoon-style stickers of Donald Trump, each showing different emotions: happy, angry, surprised, confused. Bold outlines, simple cel-shading, and a vibrant color palette. The background must be white.

指令:一組卡通風格的(de)唐納德·特朗(lang)普(pu)表情貼(tie)紙,每(mei)張(zhang)展示不同的(de)情緒:開心、生氣、驚訝(ya)、困惑。設(she)計需有粗(cu)線條輪廓、簡單的(de)賽(sai)璐(lu)璐(lu)上色(se)(se)(se)、鮮艷的(de)配(pei)色(se)(se)(se)。背景必須(xu)為白(bai)色(se)(se)(se)。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

生(sheng)成結果出來(lai)后(hou),這套貼(tie)紙幾乎(hu)可以直接拿(na)去(qu)特朗普大樓“上(shang)架”。四(si)種情緒都緊(jin)貼(tie)特朗普的(de)人設:開心時(shi)眉(mei)飛色(se)舞,生(sheng)氣時(shi)火力全開,四(si)種表情都夸張(zhang)到位。

如果想要(yao)復刻類(lei)似風格(ge),可以嘗試(shi)在提示(shi)詞里加上“cel-shading(賽璐(lu)璐(lu)上色)”,加上“cel-shading”就是告訴(su)模型:別(bie)給我太寫實的陰影(ying),要(yao)畫成(cheng)卡通風格(ge)的色塊效果,就像動畫人物那樣。

二、產品廣告篇(共4類)

產(chan)品(pin)廣告(gao)篇共有4類玩法(fa),包含明(ming)盒盲(mang)盒、商品(pin)植入、產(chan)品(pin)內(nei)部圖(tu)和廣告(gao)分(fen)鏡腳本。

8、明盒、盲盒

Prompt 1:Create a photorealistic product mockup featuring a stylized figurine of the character from the reference image, posed according to the action reference image. The figurine should stand on a display base, placed next to its packaging box decorated with original artwork. Set the scene on a white studio surface with bright, even lighting to accentuate textures and branding. Use a 50?mm product portrait lens to capture fine detail and depth. The packaging design must align with the figurine’s theme—vibrant colors, clear typography, and a glossy finish. Ensure visual and identity consistency between the figurine and reference images, including pose, style, lighting, and proportions, resulting in a cohesive commercial-quality product shot.

指令1:寫實(shi)風(feng)格的產(chan)品Mockup,展示根據(ju)參(can)考圖中(zhong)的角(jiao)色(se)制作(zuo)的手(shou)辦,并擺(bai)出(chu)與動(dong)作(zuo)參(can)考圖中(zhong)姿勢(shi)一致(zhi)的動(dong)作(zuo)。手(shou)辦置于展示底座上,旁(pang)邊放置印(yin)有原始藝術形象的包裝盒。場景設(she)定于白(bai)色(se)攝影(ying)棚(peng)臺面,采用明亮均勻的光線以(yi)突出(chu)材質質感和(he)品牌細(xi)(xi)節(jie)。使用50毫(hao)米(mi)產(chan)品肖像(xiang)鏡(jing)頭(tou)捕捉精細(xi)(xi)細(xi)(xi)節(jie)與景深。包裝設(she)計需(xu)與手(shou)辦主題保(bao)持一致(zhi),色(se)彩鮮艷、排版清晰、表面有光澤(ze)。手(shou)辦造型需(xu)嚴(yan)格對應參(can)考圖的角(jiao)色(se)設(she)定,包括(kuo)姿勢(shi)、風(feng)格、光線和(he)比例(li),整體畫(hua)面統一,表現出(chu)專業的商業產(chan)品攝影(ying)效果。

Prompt 2:Create a 1/7 scale commercial figurine of the character from the reference image, in a photorealistic style and realistic environment. The figurine is placed on a computer desk, standing on a round transparent acrylic base with no text. On the computer screen, display the ZBrush modeling process of the figurine. Next to the computer screen, place a BANDAI-style toy packaging box featuring the original artwork of the character.

指(zhi)令(ling)2:繪制圖中角色的(de)(de)1/7比例商業(ye)化手(shou)(shou)(shou)辦,寫實風格(ge),真實環境。手(shou)(shou)(shou)辦放置(zhi)在電腦桌上,使用(yong)圓形透明亞克力底座(zuo),底座(zuo)上無任何文(wen)字。電腦屏幕(mu)里顯示的(de)(de)是該手(shou)(shou)(shou)辦的(de)(de)ZBrush建模過程。在電腦屏幕(mu)旁,放置(zhi)一只(zhi)印有原畫(hua)的(de)(de)、具有BANDAI風格(ge)的(de)(de)玩(wan)具包裝盒。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參考圖(reference image)

在網絡上已經(jing)爆火的(de)盲盒款(kuan)之外,智東(dong)西還(huan)試了試“明盒款(kuan)”。生成結果(guo)里,玩具的(de)透(tou)明包裝盒細節相當(dang)逼(bi)真(zhen),手辦(ban)的(de)質感(gan)和比例也十(shi)分到位,看(kan)上去(qu)就(jiu)像是真(zhen)實上架(jia)的(de)商品(pin)照。如果(guo)不是親手跑出(chu)來,很(hen)容(rong)易以為(wei)是哪家廠(chang)商提前放出(chu)的(de)宣(xuan)傳圖。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

9、商品植入

Prompt: A sleek, photorealistic commercial image featuring the female model from the reference image holding and presenting the figurine from the other reference image. The scene should have the atmosphere of a high-end toy advertisement, with soft, natural lighting and a minimalist, lifestyle-oriented background. Use a 50?mm product portrait lens to capture sharp detail and depth. Ensure the model’s pose, lighting, and styling are consistent with her reference image, and the figurine’s proportions, textures, and colors match its reference as well. The figurine must appear naturally integrated, with accurate shadows, reflections, and scale relative to the model. The overall composition should feel cohesive, aspirational, and suitable for promotional use.

指(zhi)令:一幅精致寫實(shi)風(feng)格(ge)的(de)商(shang)業廣告圖,基于(yu)參考(kao)圖中的(de)女模特(te)(te)姿(zi)態與(yu)(yu)(yu)(yu)(yu)光(guang)(guang)線(xian),手持并(bing)展示另一個參考(kao)圖中的(de)玩(wan)偶。場(chang)景(jing)營造高端玩(wan)具(ju)廣告風(feng)格(ge),背(bei)景(jing)簡潔極簡且具(ju)生活氛圍,光(guang)(guang)線(xian)柔和(he)自(zi)然(ran),突出溫馨與(yu)(yu)(yu)(yu)(yu)時尚感。使(shi)用50mm產品肖像鏡頭拍(pai)攝,以捕捉(zhuo)細節與(yu)(yu)(yu)(yu)(yu)景(jing)深。女模特(te)(te)的(de)姿(zi)勢、光(guang)(guang)線(xian)與(yu)(yu)(yu)(yu)(yu)造型需(xu)與(yu)(yu)(yu)(yu)(yu)其(qi)參考(kao)圖保(bao)持一致;玩(wan)偶的(de)比(bi)例(li)、材質質感與(yu)(yu)(yu)(yu)(yu)色彩需(xu)與(yu)(yu)(yu)(yu)(yu)其(qi)參考(kao)圖匹配。確保(bao)玩(wan)偶與(yu)(yu)(yu)(yu)(yu)模特(te)(te)融合自(zi)然(ran),陰影、反光(guang)(guang)與(yu)(yu)(yu)(yu)(yu)尺(chi)寸(cun)比(bi)例(li)恰當。畫面整體(ti)統一協調,具(ju)備宣傳級質感。

在這組效(xiao)果里,模特(te)原本手(shou)里拿著的(de)是(shi)一塊蛋糕,被Nano Banana“無縫”替換(huan)成了LABUBU玩偶。手(shou)部的(de)銜(xian)接尤其值得(de)一提(ti),手(shou)指彎曲的(de)角度和(he)指尖(jian)都處理(li)得(de)很細致,沒有出現常見的(de)“手(shou)指畸(ji)形(xing)”。與此同時,玩偶毛(mao)茸茸的(de)質(zhi)感也(ye)被完整(zheng)保留下來,和(he)模特(te)動(dong)作自然融(rong)合,整(zheng)張圖看上去(qu)就像一張真的(de)玩具廣告照。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

10、產品內部圖

Prompt: A hyper-realistic exploded-view photograph of a refrigerator, showing only the user-visible parts such as shelves, drawers, doors, and compartments. Each part is separated and floating apart in a controlled explosion layout, as if suspended in mid-air. The lighting is soft and even, highlighting realistic materials such as glass, plastic, and metal surfaces. All components must remain in accurate scale and perspective. The refrigerator must be fully complete with no missing parts. The background should be clean white or neutral, creating the look of a professional product display.

指令:一幅(fu)高(gao)度寫(xie)實的冰箱爆(bao)炸視圖照片,僅展示(shi)用戶(hu)可(ke)見的部(bu)(bu)分(fen),如擱(ge)板、抽屜、門體(ti)和(he)(he)儲物格。各個部(bu)(bu)件以“controlled explosion(可(ke)控爆(bao)炸視圖)”的方式分(fen)離(li)懸浮(fu),仿佛漂浮(fu)在空(kong)中。光線柔和(he)(he)均(jun)勻,突(tu)出玻璃、塑料和(he)(he)金屬表面的真實質感。所有部(bu)(bu)件需保持比例與透(tou)視準確。冰箱必須完整呈現(xian),不能缺少(shao)任何部(bu)(bu)件。背景為干凈的白色(se)或中性色(se),呈現(xian)專(zhuan)業產品展示(shi)的效果。

Nano Banana生成的(de)畫面里(li),冰箱的(de)擱板、抽屜(ti)、門體都(dou)被整齊地分離開,材(cai)質質感(gan)清晰,還原得(de)很逼真(zhen)。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

11、廣告分鏡腳本

Prompt:A four-frame storyboard for a new running shoe commercial in a sports event montage style. The sequence should feel like dynamic highlights from an intense competition, with photorealistic detail. Strong lighting contrasts, motion blur, and heavy trailing effects. Each frame should capture energetic moments of movement and intensity, leading to a final product-focused shot with the shoe and brand logo.

指令:一組四格分鏡腳(jiao)本,用于新跑鞋廣(guang)告,以體(ti)育賽(sai)事蒙太奇風格呈現(xian)。整體(ti)氛圍如同激烈(lie)比(bi)賽(sai)的動(dong)態集錦,細節寫實。畫面中(zhong)光(guang)影(ying)對(dui)比(bi)強(qiang)烈(lie),運動(dong)模糊(hu)與拖影(ying)效果加重。每(mei)一格都展現(xian)充滿能量與張力的動(dong)態瞬間,最(zui)后結(jie)尾于以跑鞋和品(pin)牌(pai)Logo為(wei)核心的產品(pin)鏡頭(tou)。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參考圖(reference image)

每(mei)一格畫面都準確捕捉了運動瞬(shun)間,最后一格也(ye)自(zi)然過(guo)渡到產品(pin)展示,跑鞋與品(pin)牌Logo完(wan)整呈現,畫面干凈利落。值(zhi)得一提的是,英文Logo與文字部分也(ye)很穩定,沒有出(chu)現常(chang)見的亂碼問題。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

三、美術視覺篇(共7類)

美(mei)術視覺(jue)篇共(gong)有7類玩(wan)法,包含換風格(ge)換角度、線稿上(shang)色(se)、打光、生(sheng)成連環畫、寶可夢卡片(pian)、游戲UI和多元素拼接。

12、換種風格、換個角度

Prompt 1: Recreate the reference image in a bright and cheerful style, with vivid colors, lively atmosphere, and enhanced vibrancy. The overall mood should feel joyful and uplifting, while keeping the original subject recognizable.

指令1:將(jiang)參考圖重(zhong)新創(chuang)作成明(ming)(ming)艷(yan)歡快的風格(ge),畫(hua)面(mian)色彩鮮明(ming)(ming),氛圍活潑(po),整體更具亮度與生機。在(zai)保(bao)持原始(shi)主體可(ke)識別的同時(shi),營造出愉(yu)悅、積極的觀(guan)感。

智東西把(ba)參考圖中的(de)“黑(hei)化版”懶洋(yang)洋(yang)交給(gei)Nano Banana處理(li),結果(guo)立(li)刻“洗白”成了本色的(de)蠢(chun)萌(meng)路線。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

Prompt 2: Recreate the reference image from multiple different camera angles—such as front view, top-down view, low-angle view, and 45-degree side view—while keeping the subject recognizable. Arrange all variations together within a single image on a clean white background, ensuring consistency in style and lighting.

指令2:將參(can)考圖以多個(ge)不(bu)(bu)同(tong)的拍攝視(shi)(shi)(shi)(shi)角(jiao)(jiao)重新創作(zuo),如正面視(shi)(shi)(shi)(shi)角(jiao)(jiao)、俯視(shi)(shi)(shi)(shi)角(jiao)(jiao)、仰視(shi)(shi)(shi)(shi)角(jiao)(jiao)和45度側視(shi)(shi)(shi)(shi)角(jiao)(jiao),在保持主(zhu)體可識別(bie)的同(tong)時展現差異。將所有不(bu)(bu)同(tong)視(shi)(shi)(shi)(shi)角(jiao)(jiao)的畫面排布在同(tong)一張(zhang)白色背(bei)景的圖中,保證風格與光線(xian)一致。

智(zhi)東西(xi)選用了(le)詹姆(mu)斯中(zhong)國行時的(de)一張(zhang)上海(hai)游客照,交給Nano Banana去生成多個角度的(de)版本。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參(can)考圖(reference image)

結果不(bu)僅(jin)人物在不(bu)同視角(jiao)下保(bao)持了高(gao)度一(yi)致性,還自然(ran)衍生出(chu)幾種合(he)影姿勢(shi),看(kan)起(qi)來十(shi)分寫實。細節上,他手里拿著(zhu)的籃(lan)球(qiu)也會隨著(zhu)視角(jiao)變化而(er)調整(zheng)位置和(he)角(jiao)度,整(zheng)體效果就像(xiang)是一(yi)組真(zhen)實拍下的多角(jiao)度照片。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

13、線稿上色

Prompt: Colorize an action line drawing featuring two characters. The action poses and composition must remain exactly the same as in the line art, while the characters’ appearances, proportions, and details must strictly match the reference images. Apply colors in the same style as the references, with balanced tones, clear highlights, and shadows. Keep the line work sharp and visible. Add a dynamic background in the style of Japanese anime battle scenes—fiery, high-energy, and dramatic—with glowing effects, speed lines, or explosive lighting to enhance the intensity of the action. Ensure that the anatomy and body structure of the characters are realistic and proportional.

指令:為一(yi)張包含兩(liang)個人(ren)物的動作線稿(gao)進行上色。動作姿勢與線稿(gao)圖(tu)保持完全一(yi)致,人(ren)物的外(wai)觀、比(bi)例和(he)(he)細節需嚴格(ge)遵(zun)循參考圖(tu)。上色風格(ge)需與參考圖(tu)統一(yi),色彩(cai)均衡(heng),明(ming)暗和(he)(he)高(gao)光(guang)清晰。保證線條清晰可(ke)見(jian)。背景添加(jia)日(ri)本動漫熱血戰斗風格(ge),充(chong)滿燃感和(he)(he)張力,可(ke)以(yi)加(jia)入(ru)光(guang)效、速度線或爆炸般的光(guang)影效果(guo),以(yi)強化(hua)動作的激烈氛圍。同時(shi)要保證人(ren)物的人(ren)體構造合(he)理,比(bi)例自然。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參考(kao)圖(reference image)

網上的(de)部分案例中,Nano Banana的(de)線稿上色(se)效果相當成(cheng)(cheng)熟(shu),生成(cheng)(cheng)結果都能做到細致填色(se)。智東西還(huan)希望Nano Banana不(bu)僅(jin)能上色(se),還(huan)要能加上角色(se)形(xing)象和日漫式(shi)戰(zhan)斗氛圍。

起初,當輸入一張元素(su)較(jiao)為(wei)復雜(za)的(de)線稿時,生(sheng)成結果在部分元素(su)的(de)處理上出(chu)現了偏差。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲復雜版線稿生成(cheng)效果

但在換成結(jie)構更清晰的(de)線稿后,Nano Banana就順(shun)利完成了(le)任務,角色(se)細節到位(wei),背(bei)景里也加上了(le)速度線、光(guang)效(xiao)等典(dian)型(xing)的(de)熱血元素(su),看起來頗有(you)一股動漫原(yuan)畫(hua)的(de)氣勢。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

14、打光

Prompt: A photorealistic portrait illuminated by a strong frontal spotlight. The direct light clearly reveals the subject’s features and textures, while casting a sharp, dramatic shadow onto the wall behind them. Ensure the subject is positioned at a certain distance from the wall so that the shadow falls clearly onto it. The atmosphere should feel intense and theatrical, with the contrast between the brightly lit subject and the deep shadow on the wall enhancing the visual impact.

指令:一幅寫實風格的(de)人(ren)像作品(pin),由正面(mian)(mian)的(de)一束強烈聚光(guang)燈照亮(liang)。正面(mian)(mian)光(guang)清晰(xi)展現人(ren)物的(de)特征與質感(gan),同時在背后的(de)墻(qiang)面(mian)(mian)上(shang)投(tou)射出戲劇化(hua)(hua)的(de)影(ying)子(zi)。要保證人(ren)物與墻(qiang)面(mian)(mian)之(zhi)間(jian)留有一定距離,使影(ying)子(zi)能夠清晰(xi)地落在墻(qiang)面(mian)(mian)上(shang)。整(zheng)體氛圍應充滿張力(li)與舞臺(tai)氛圍,通過明亮(liang)的(de)人(ren)物與深(shen)邃的(de)墻(qiang)面(mian)(mian)陰(yin)影(ying)之(zhi)間(jian)的(de)對(dui)比,強化(hua)(hua)視(shi)覺沖擊(ji)力(li)。

智(zhi)東西(xi)選用了一張迪(di)麗(li)熱巴近(jin)期“火出圈”的一張紅毯(tan)出發圖,生(sheng)成(cheng)結果(guo)里(li),正面(mian)光不僅準確地(di)給(gei)出了人(ren)物的動作,墻(qiang)面(mian)上的投影也極為逼真。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

15、連環畫

Prompt:A sequence of three comic panels in a colorful superhero style featuring Donald Trump and Captain America in conversation:

(1) Captain America lands dramatically on a rooftop where Trump is waiting. Trump says in a bold dialogue box: “You think you can protect everyone?”

(2) The two face each other, engaging in tense dialogue. Captain America replies: “That’s my duty—and I won’t back down.”

(3) A wide shot shows them both in dynamic poses, their expressions intense. A bold caption box at the bottom reads: “The clash of ideals begins!”

Use vibrant colors, bold outlines, and classic comic book style with clear speech bubbles.

指令(ling):一組三(san)格(ge)漫畫,采用色彩鮮艷的超(chao)級英雄風格(ge),主角為唐納德·特朗普和(he)美國隊長的對話:

(1)美國隊長(chang)戲(xi)劇(ju)性地降(jiang)落(luo)在(zai)特朗(lang)普等待的屋(wu)頂(ding)上。特朗(lang)普在(zai)對白(bai)框中說:“你以為(wei)你能保護所有人嗎(ma)?”

(2)兩人(ren)面(mian)對(dui)面(mian)展開(kai)緊張的對(dui)話。美國(guo)隊長回應:“那是我的職(zhi)責,我絕不會退(tui)縮(suo)。”

(3)遠景(jing)展(zhan)現兩人充(chong)滿(man)張力的(de)動態(tai)姿勢,神情激烈。畫面下方的(de)旁(pang)白框寫著:“理念(nian)的(de)碰撞,正(zheng)式開始!”

畫面采用(yong)鮮明(ming)的色彩、粗獷的線條和經(jing)典漫(man)畫風格,配以清晰(xi)的對(dui)白氣泡。

在這(zhe)一組漫(man)畫(hua)分鏡中,Nano Banana雖然準(zhun)確生成了(le)超級(ji)英雄風格的(de)畫(hua)風、對(dui)白(bai)框和旁白(bai)框,但還是(shi)出現了(le)典型的(de)“小翻車”:同一畫(hua)格里跑出了(le)兩(liang)個美國(guo)隊(dui)長(chang)、兩(liang)個特朗(lang)普(pu)。好在整體(ti)氛圍和細節都在線(xian),看上去已經很接(jie)近一張標準(zhun)的(de)漫(man)畫(hua)分鏡稿。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

16、寶可夢卡片

Prompt: A custom Pokémon card generated based on the reference image. The card design should strictly follow the style, layout, and details of the reference, while keeping the overall look consistent with the classic Pokémon TCG style. Ensure the character, stats, and abilities are adapted from the reference image, and the final card feels authentic and collectible.

指令:根據參考圖(tu)(tu)生成一(yi)張自(zi)定(ding)義寶(bao)可夢卡牌。卡面設計需嚴(yan)格遵循參考圖(tu)(tu)的風(feng)格、排版與細節,同時(shi)整體效果保持與經典Pokémon TCG風(feng)格一(yi)致。角色形象、數值(zhi)和技能需基(ji)于參考圖(tu)(tu)進行改編(bian),最(zui)終呈現(xian)出真實(shi)且具(ju)收藏(zang)感的卡牌。

Nano Banana直接把扎克伯格生成進了寶可夢(meng)卡(ka)牌(pai)里。最妙的(de)是,它(ta)還(huan)很“懂人設”,自動(dong)給小(xiao)札分配的(de)技能(neng)是“好友請(qing)求”和(he)(he)“數據”。畫面風(feng)格和(he)(he)經(jing)典Pokémon TCG完全對齊。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

17、游戲UI

Prompt:A game UI mockup generated based on the reference image. The layout, color scheme, and styling should strictly follow the reference, while maintaining the look and feel of a functional game interface. Include core elements such as health bars, inventory slots, and a quest log, all adapted to match the reference’s design language. Ensure the final UI feels polished, cohesive, and visually consistent with the reference style.

指令:根據參(can)(can)考(kao)(kao)(kao)圖(tu)(tu)生(sheng)成一(yi)份游戲(xi)UI界面(mian)模型。界面(mian)布局、配色(se)和風格需(xu)嚴格遵循參(can)(can)考(kao)(kao)(kao)圖(tu)(tu),同時保證(zheng)具備完整的(de)游戲(xi)交(jiao)互感。包含生(sheng)命值條、物(wu)品欄(lan)和任(ren)務日志等核心(xin)元(yuan)素(su),并根據參(can)(can)考(kao)(kao)(kao)圖(tu)(tu)的(de)設計語言進行改編(bian)。最(zui)終的(de)UI需(xu)精致統一(yi),視覺效(xiao)果(guo)與參(can)(can)考(kao)(kao)(kao)風格保持一(yi)致。

生(sheng)(sheng)成的(de)(de)效果第一眼(yan)看上(shang)去,就(jiu)像是老式那(nei)種經典卡牌游戲界面:生(sheng)(sheng)命值條、物品欄、任務日志一應俱全,配色也帶著一點復古味(wei)道。雖(sui)然(ran)是AI生(sheng)(sheng)成的(de)(de)UI mockup,但(dan)整體精致(zhi)統(tong)一,有“能直接玩起來”的(de)(de)感覺。智東西后續有嘗試生(sheng)(sheng)成塞爾達和原神這(zhe)類大世界探索(suo)游戲的(de)(de)UI設計,但(dan)Nano Banana無法理解。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

18、多元素拼接

Prompt: Compose a full-body photorealistic image of the model from one reference image, wearing all the clothing, shoes, and accessories from the other reference image. Ensure every item is transferred accurately, including proportions, textures, colors, and fine details, and that the entire outfit fits the model’s body and pose naturally. The final result should look seamless and realistic, as if the model is truly wearing the complete set of garments, footwear, and accessories from the clothing reference.

指令:生(sheng)成一(yi)張模(mo)特(te)的全(quan)身寫實合成圖,以一(yi)張參考(kao)圖中的模(mo)特(te)為(wei)主體,并(bing)讓(rang)其穿上另一(yi)張參考(kao)圖中的所有衣服、鞋(xie)子和配飾。需完整(zheng)(zheng)呈現(xian)每一(yi)件單品,保持比例、材質、顏色和細節的準(zhun)確(que)性,并(bing)確(que)保整(zheng)(zheng)體穿搭(da)自然(ran)(ran)貼合模(mo)特(te)的體態與姿勢。最終效果應無縫融(rong)合,真實自然(ran)(ran),就(jiu)像模(mo)特(te)本就(jiu)穿著參考(kao)圖里的完整(zheng)(zheng)服裝、鞋(xie)履(lv)和配飾一(yi)樣。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參考(kao)圖(reference image)

在這次測試中(zhong),服飾的每個細(xi)節都被完(wan)整保留下來:布料(liao)的材(cai)質和(he)配飾的顏色都精準復現。生成的畫面幾乎沒有違和(he)感(gan),看起來就像模特真的穿(chuan)上(shang)了(le)另一張(zhang)圖里的全(quan)套穿(chuan)搭。效(xiao)果相(xiang)當驚(jing)艷,讓(rang)人(ren)感(gan)覺Nano Banana完(wan)全(quan)可以勝任“虛擬試衣間”的角色。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

四、空間合成篇(共2類)

空間合成篇(pian)共有(you)2類玩(wan)法,包含3D建筑(zhu)建模和地(di)標AR圖(tu)。

19、3D建筑建模

Prompt:A detailed 3D model of the building from the reference image, presented with dashed lines to highlight structural outlines and hidden sections. The model should maintain realistic proportions and architectural accuracy, while the dashed lines clearly illustrate the building’s form and internal layout.

指令:根(gen)據(ju)參考圖生成該建筑(zhu)(zhu)的(de)精(jing)細3D模(mo)型,以虛線方(fang)式(shi)呈現結(jie)構輪(lun)廓和隱藏部分。模(mo)型需(xu)保持真實(shi)比例與建筑(zhu)(zhu)細節的(de)準確性,同時通過(guo)虛線清晰展示建筑(zhu)(zhu)的(de)形態與內部布局。

智東西選取了(le)巴塞羅那大教(jiao)堂作為測試對象。生成結果(guo)不(bu)僅給出了(le)45度俯視角的(de)效果(guo),還生成了(le)一張完(wan)整的(de)立體模(mo)型圖。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

20、地標AR圖

Prompt:A photorealistic photo of the landmark from the reference image, with AR overlays highlighting historical facts, structural details, measurements, and visitor information. The overlays should look like an augmented reality interface, blending naturally with the real-world photo while remaining clear and easy to read.

指令:根(gen)據(ju)參考圖(tu)生成該地標的寫實(shi)照片,并添加AR覆蓋(gai)層,顯示(shi)歷史信息、結構(gou)細(xi)節、尺寸測量(liang)和游(you)客指南(nan)。覆蓋(gai)層需具備AR的效果,與真實(shi)照片自(zi)然融合(he),同(tong)時保持清晰易讀。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

▲參考(kao)圖(reference image)

智東西(xi)選擇了上海的(de)(de)陸家嘴(zui)作為生成(cheng)對(dui)象。結果畫面在保持真實攝影風格的(de)(de)同(tong)時,疊(die)加了信息層(ceng):包(bao)括(kuo)高度、建成(cheng)年份、功(gong)能分區等數據,還有(you)游(you)客可(ke)見的(de)(de)觀景臺提示(shi)。整體效果像(xiang)是手機(ji)舉起(qi)來(lai)時的(de)(de)AR導(dao)航界面,信息點(dian)和建筑自然對(dui)應,讀起(qi)來(lai)直觀清(qing)晰。

Nano Banana“終極玩法指南”上線!從合影到3D建模,AI作圖“全能王”誕生

就這輪20種玩法來(lai)看(kan),Nano Banana在寫實合成、產品級(ji)渲染和多視角一致性(xing)(xing)上(shang)驚喜(xi)不(bu)少,偶爾也會在復雜元(yuan)素與(yu)分鏡里“翻(fan)車”,但整體可玩性(xing)(xing)與(yu)可用度(du)都已足夠高。

不過(guo),隨著模(mo)型繼續迭代,未來的驚喜也值得期(qi)待:比如(ru)更(geng)好的人(ren)物一致性、更(geng)自然的手(shou)部和表情生成、更(geng)懂語(yu)境的分鏡(jing)連貫性,甚至可能支持更(geng)長鏈(lian)路的創作(zuo)流程,讓一次指令就能完成“成品”。

到那時,AI或將真正變成創(chuang)作者的(de)常用工具。