# Whisk AI 图片混合器：融合主体、场景与 风格

> Whisk AI 把三张图变成一张：放入主体、场景和风格，愿意的话再写几个字，就能用 Nano Banana 或 GPT Image 得到一张全新的图片，还能一键做成短视频。

混合工作室: https://whiskai.net/zh/image · 做成视频: https://whiskai.net/zh/video · 价格: https://whiskai.net/zh/pricing · 画廊: https://whiskai.net/zh/gallery

WhiskAI 是独立服务，与 Google 无关联，也未获其认可或赞助；Whisk 和 Google 是 Google LLC 的商标。

## 什么是 Whisk AI？

Whisk AI（whiskai.net 上的 WhiskAI）是一个在线图片混合工具。把最多三张图片放进带标签的主体、场景和风格槽位，再加一句可选的提示词，Nano Banana 或 GPT Image 就会把它们融合成一张新图片。任何结果都能用 Seedance、Veo 3.1、Kling 4 或 Grok Imagine 做成短视频。

一个积分余额通用于所有模型。生成前就能看到准确的消耗，失败会自动退还积分。WhiskAI 是独立服务，与 Google 无关联，也未获其认可或赞助；Whisk 和 Google 是 Google LLC 的商标。

## Whisk AI 如何工作：主体、场景、风格

每个槽位告诉模型要从那张图里取什么。可以只放一张、两张或三张；提示词可写可不写。

- **主体** — 你想保留的人物、宠物、产品或物体。Whisk AI 会保留它的样子、外形和关键细节。 适合放： 主体清晰的单张照片、吉祥物草图、白底产品图。
- **场景** — 主体要放到哪里：房间、街道、风景或版式。场景决定背景和构图。 适合放： 旅行照片、室内、货架、海报版式。
- **风格** — 结果要呈现的样子：媒介、配色、质感和光线。只借用风格图的“样子”，不借用它的内容。 适合放： 水彩画、黏土渲染、孔版印刷、贴纸页。
- **一张新图** — Whisk AI 按顺序把槽位图片连同角色说明和你的提示词发给 Nano Banana 或 GPT Image，得到的是一张全新的图，而不是拼贴。

没有图片？Whisk AI 也能当普通的文生图工具：槽位留空，直接描述画面即可。

## 用 Whisk AI 三步完成混合

1. **放入图片** — 放入主体、场景和风格图片（任选，均可不放）。每张支持 JPG、PNG 或 WebP，最大 10 MB。
2. **加一句要求，选模型** — 写一句简短要求，如“做成珐琅徽章”，也可以留空。追求还原度选 Nano Banana Pro，想要大胆的图形感选 GPT Image，再选尺寸和比例。
3. **混合、下载、做成视频** — 大多数混合一分钟内完成。可以下载、复用设置做变体，或点“做成视频”生成短片。

## Whisk AI 风格灵感：贴纸、手办、毛绒、海报

常见的 Whisk AI 混合风格，每种都附一条可直接粘贴的提示词公式，把方括号里的词换成你自己的主体即可。

- **模切贴纸** — 粗描边、白边、平涂，适合电脑贴纸和聊天表情包。 `[主体] 做成亮面模切贴纸，粗白边，粗描边，平涂配色，居中放在纯色背景上`
- **毛绒玩具** — 柔软布料、缝线和圆润比例，把任何角色变成可爱的玩偶。 `[主体] 做成柔软的毛绒玩具，毛毡和短绒布料，可见缝线，圆润比例，粉彩架子上的棚拍照片`
- **珐琅徽章** — 金属勾边、亮面珐琅填色加上背卡，一看就是周边商品。 `[主体] 做成硬珐琅徽章，金色金属勾边，亮面填色，别在牛皮纸背卡上，微距产品照`
- **3D 手办** — 底座上的收藏级搪胶手办，柔和棚光，带点玩具盒的感觉。 `[主体] 做成圆形底座上的收藏级搪胶手办，哑光塑料质感，柔和棚光，背景是玩具包装盒`
- **黏土风** — 手捏橡皮泥质感，带指纹和暖光，像定格动画的场景。 `[主体] 用橡皮泥捏成，可见指纹，手工质感，温暖柔光，迷你定格动画场景`
- **海报** — 大字标题、强烈图形和有限配色，适合演出海报和活动传单。 `[主体] 放在大胆的丝网印刷海报上，限定三色，大标题“[文字]”，颗粒质感，居中版式`
- **水彩** — 松散的晕染、纸张纹理和柔和边缘，适合绘本和贺卡风格。 `[主体] 画成松散的水彩插画，柔和晕染，可见纸纹，边缘轻微渗色，留白边`
- **产品图** — 把你的产品放进新场景，干净的商业布光，适合商品页和广告。 `[产品] 放在[场景]中，干净的商业布光，柔和阴影，浅景深，电商主图`

## Whisk AI 里的 Nano Banana：忠实的多图混合

Google 的 Nano Banana 系列是 Whisk AI 的默认引擎。它能同时读取多张参考图，保持主体可辨认，并把它自然地放进新场景。Nano Banana Pro 还提供棚拍级光线、清晰文字和 4K 输出。

## Whisk AI 里的 GPT Image：大胆的图形化混合

OpenAI 的 GPT Image 模型擅长遵循长指令，偏爱图形化风格：海报、贴纸、包装和大量文字的版式。当风格槽位比照片真实感更重要时，用它最合适。

- **Nano Banana Pro** — Google’s flagship — sharp text, studio lighting and up to 4K. 1K / 2K / 4K, up to 8 reference images.
- **Nano Banana 2.1** — Fast, faithful edits and character consistency at a low price. 1K / 2K / 4K, up to 8 reference images.
- **Nano Banana 2** — Detailed renders with wide aspect ratios and 4K output. 1K / 2K / 4K, up to 8 reference images.
- **Nano Banana 2 Lite** — The quickest Nano Banana for drafts and variations. 1K, up to 8 reference images.
- **Nano Banana** — The original Gemini image model — great for edits. 1K, up to 8 reference images.
- **GPT Image 2.5 Flare** — OpenAI’s newest — vivid, stylized images with crisp typography. 1K / 2K / 4K, up to 8 reference images.
- **GPT Image 2.5 Sunburst** — Warm, photographic GPT Image 2.5 tuned for realism. 1K / 2K / 4K, up to 8 reference images.
- **GPT Image 2** — Reliable layouts, posters and product shots up to 4K. 1K / 2K / 4K, up to 8 reference images.
- **GPT Image 2.6 Economy** — Low-cost GPT Image 2.6 for quick text-to-image ideas. 1K, up to 0 reference images.
- **GPT Image 1.5** — Medium or high quality GPT Image with strong prompt following. 1K / 2K, up to 16 reference images.
- **GPT Image 1 (4o)** — The classic GPT-4o image model — friendly, illustrative results. 1K, up to 5 reference images.

## 把 Whisk AI 混合图做成视频

每张生成好的图片都有“做成视频”按钮。它会打开视频工作室，并把你的混合图设为首帧，你只需描述动作：镜头缓慢推进、挥手、转头。

- **Veo 3.1 Fast** — Google’s cinematic model with native audio — fast renders up to 4K. 4–8s, 720p / 1080p / 4K, aspect ratios 16:9, 9:16.
- **Veo 3.1 Quality** — Highest-fidelity Veo 3.1 for hero shots, with dialogue and sound. 4–8s, 720p / 1080p / 4K, aspect ratios 16:9, 9:16.
- **Veo 3.1 Lite** — The most affordable Veo 3.1 — great for drafts and social clips. 4–8s, 720p / 1080p / 4K, aspect ratios 16:9, 9:16.
- **Seedance 2.5** — ByteDance’s newest — up to 30s, 30 reference images and 4K. 4–30s, 480p / 720p / 1080p / 4K, aspect ratios 16:9, 9:16, 1:1, 4:3, 3:4, 21:9.
- **Seedance 2.5 Flash** — Faster, cheaper Seedance 2.5 at 480p or 720p. 4–30s, 480p / 720p, aspect ratios 16:9, 9:16, 1:1, 4:3, 3:4, 21:9.
- **Seedance 2.0** — Multi-shot storytelling with native audio, up to 15s and 4K. 4–15s, 480p / 720p / 1080p / 4K, aspect ratios 16:9, 9:16, 1:1, 4:3, 3:4, 21:9.
- **Seedance 2.0 Fast** — Quicker Seedance 2.0 renders at 480p or 720p. 4–15s, 480p / 720p, aspect ratios 16:9, 9:16, 1:1, 4:3, 3:4, 21:9.
- **Seedance 2.0 Mini** — The lowest-cost Seedance model — ideal for testing ideas. 4–15s, 480p / 720p, aspect ratios 16:9, 9:16, 1:1, 4:3, 3:4, 21:9.
- **Kling 4.0** — Up to 30s with native audio, 10 keyframes and omni reference. 3–30s, 720p / 1080p / 4K, aspect ratios 16:9, 9:16, 1:1, 21:9, 4:3, 3:4.
- **Kling 4.0 Flash** — Faster, lower-cost Kling 4 drafts — up to 20s at 720p. 3–20s, 720p, aspect ratios 16:9, 9:16, 1:1, 21:9, 4:3, 3:4.
- **Grok Imagine** — xAI’s video model with sound — 6 to 30 seconds from text or images. 6–30s, 480p / 720p / 1080p, aspect ratios 16:9, 9:16, 1:1, 2:3, 3:2.

## Whisk AI 工具箱里有什么

- **带标签的混合槽位** — 主体、场景、风格各有独立槽位，部分模型总共可接收最多 16 张参考图。
- **提示词可选** — 放好图片直接点“生成”。文字能帮你引导结果，但不是必需的。
- **Nano Banana 和 GPT Image** — 每次都可以选择 Google 的 Nano Banana 系列或 OpenAI 的 GPT Image 系列来混合。
- **最高 4K** — 先用 1K 出草稿，满意后在支持的模型上以 2K 或 4K 出图。
- **任意比例** — 贴纸用方图，信息流用 4:5，快拍用 9:16，横幅用 16:9，也可以自动。
- **复用与重跑** — 每个结果都会记住槽位、提示词和模型，一键复用即可做变体。
- **做成视频** — 用 Seedance、Veo 3.1、Kling 4 或 Grok Imagine 把混合好的图做成短视频。
- **按张计费** — 一个积分余额通用，图片 6 积分起，失败自动退还积分。
- **私密模式** — 付费套餐可让混合结果不进入公开画廊。上传的图片只用于你自己的生成。

## Whisk AI 在各行业的用法

- **周边与按需印刷** — 下单打样之前，先把吉祥物或宠物做成贴纸、徽章和毛绒样机图。
- **电商商品页** — 把一张产品照放进不同季节的场景和风格里，用于商品页、广告和邮件。
- **社交媒体** — 把自拍混合成海报、黏土小人或水彩，用于信息流和快拍。
- **角色与游戏** — 同一个角色保持一致，试不同服装、世界观和画风。
- **营销与活动** — 几分钟内做出符合品牌风格的活动视觉、传单和横幅样稿。
- **课堂与讲故事** — 把主角、场景和绘本风格组合起来，为故事和课程配图。

## 创作者用 Whisk AI 做的作品

- **Seedance 2.0** (video, by @aimikoda) — A painterly animated romantic drama follows a young courier and a young mechanic racing through a decaying galactic port during their final minutes together. The Courier wears a weathered crimson travel coat and carries one small flight bag; the Mechanic wears a dark work jacket marked by luminous machine dust. Keep their faces, clothing, and bag consistent throughout. Use an Arcane-inspired visual style: sculpted 3D characters finished with expressive hand-painted textures, visible brushwork, graphic silhouettes, fractured colored light, atmospheric depth, painterly motion smears, selective f…
- **Seedance 2.0** (video, by @user_589415fe) — 4 cute clay 3D characters together, white background, fast cute story in 8 seconds: 0-2s group bouncy hug (boing pop), 2-3s look at blank sign sad tilt head (bloop sad), 3-6s fast montage all working together funny - sewing, hanging lights, kicking balloons, stacking books (whoosh boop patter pop fast), 6-8.5s big celebration confetti dance jump together happy ending, loop, soft claymation, pastel colors, no text, no speech, no dialogueSound: boing pop, sad bloop, ding idea, fast boop whoosh patter, celebration popper and cheerful blips, 8 seconds, energetic
- **Seedance 2.0** (video, by @user_589415fe) — YouTube Shorts 8 seconds, vertical 9:16, 4 cute clay 3D characters: pink beanie pink sweater, purple beanie colorful vest, pink pom beanie blue hair orange vest, lavender beanie backpack. Fast slapstick football comedy, white background, soft claymation.Timeline:0-2s: pink sweater kicks football hard (kick whoosh)2-4s: blue hair character runs funny to catch, steps, comically slips and slides long across floor (squeak + long whoooosh slide)4-6s: ball bounces off invisible wall, boomerangs back fast (boing boing whoosh)6-8s: ball bonks softly on his head (soft bonk boing), lands as hat, other 3…
- **Seedance 2.0** (video, by @TechieBySA) — “A cinematic 15-second video following a tiny needle-felted wool chibi figurine — dark tousled hair, natural skin tone with no blush, brown eyes, light blue “TechieSA” hoodie, blue pants, blue sneakers — through a cozy morning coffee routine. The entire world around him is miniature scale, photorealistic needle-felted wool texture throughout. [0:00–0:02] Open on a tight close-up of a tiny silver espresso machine sitting on a miniature wooden kitchen counter. Soft morning sunlight streams in from a small window to the left, casting a warm golden glow. Thin wisps of steam curl upward in the stil…
- **Seedance 2.0** (video, by @user_ead132f2) — Ultra-realistic cinematic commercial Premium food advertising quality. 4K HDR Photorealistic rendering. Extremely high detail. Physically accurate materials and lighting. Smooth cinematic motion. Stable object consistency. Fantasy grounded in realistic physics. Duration: 10 seconds. Aspect ratio: 16:9 Modern minimalist kitchen with a clean wooden table. Bright natural daylight enters through large windows. Warm neutral color palette. Soft shadows. Consistent lighting throughout the sequence. The table is clean except for cooking ingredients.Main character: young woman the character wearing a s…
- **Seedance 2.0** (video, by @abxxai) — Create a director-style PREVIS action storyboard focused on 8 continuous action frames using only minimal stick figures. No realistic anatomy, clothing, pants, shoes, or character details. Show only movement, camera choreography, momentum trails, and ink motion, with only a small amount of necessary text.” SUBJECTS A hiphop dancer wearing a loose white top and dark oversized pants with irregular ink erosion along the fabric edges; holding an oversized heavy ink brush in the right hand with dense wet ink; the brush is always part of the movement, generating visible ink trails and ink masses wit…
- **Seedance 2.0** (video, by @0xemotions_) — NAME: ISABELLA ROLE: HEAD CHEF CORE MOOD: Confident. Precise. In total command of her kitchen. VISUAL SIGNATURE: Every motion economical — nothing wasted, nothing rushed. Painterly digital illustration character sheet — soft angular faceted shading with gentle rounded plane transitions, warm directional lighting, fine confident linework, semi-realistic proportions. Plain neutral background. Woman, early 30s, olive skin, dark hair pulled into a sleek low bun, sharp focused eyes, fitted white chef's double-breasted jacket with sleeves rolled to the forearm, black apron, small gold hoop earrings.…
- **Seedance 2.0** (video, by @azed_ai) — 3D animated cinematic short, high-quality stylized realism. No BGM, no dialogue, no subtitles, no text, with ambient sounds. # Constraints Vary the framing and angles in every cut to avoid monotony. Focus on glass, liquid, steam, glowing ingredients, hand gestures, and cozy magical detail. Keep the magic subtle and elegant, not chaotic. # Character Young potion maker. Small round glasses, messy short hair, oversized sweater, ink-stained fingers. Curious, careful, slightly nervous, but deeply focused. # Setting A tiny potion shop at night. Wooden shelves filled with bottles, dried herbs, candle…
- **Seedance 2.0** (video, by @Weeleey6) — Woman = a young East Asian woman around 25, soft elegant features, shoulder-length dark hair, a gentle emotional face, wearing a cozy beige coat and a scarf. Tender and heartfelt. Appearance only. Man = a young East Asian man around 27, soft handsome features, neat dark hair, wearing a dark wool coat. Quietly emotional, restrained. Appearance only. Sreet = a quiet city street at night in light rain, rendered photorealistically — glistening wet pavement, warm glowing streetlights and shop signs, soft bokeh of city lights, a moody romantic atmosphere, umbrellas and reflections. Warm-and-cool cin…
- **Seedance 2.0** (video, by @Dheepanratnam) — Four-panel character model sheet in one row, evenly spaced, identical original character in every panel: panel 1 full-body front view standing upright, arms relaxed, both feet visible, not cropped; panel 2 full-body back view showing the plain yellow paper back with the glossy adhesive strip along the top edge, not cropped; panel 3 tight close-up of the face with a calm neutral expression; panel 4 tight close-up of the same character screaming in panic, mouth a wide scribbled open oval with jagged scribbled teeth, eyes huge spiral scribbles. Plain flat light-grey seamless background, soft even…
- **Seedance 2.0** (video, by @aiwithlumi) — Cinematic Wuxia sequence in an autumn maple forest: two martial artists, a female warrior in flowing pink-and-white silk Hanfu and a male warrior in blue robes, face off on wet mossy river rocks as golden sunbeams pierce mist and red maple leaves fall. She leaps into the air, draws her steel Jian, and glides toward him in dramatic slow motion before their blades clash with a burst of glowing sparks. She lands gracefully, splashing water around her, while a black sumi-e ink-wash effect spreads across the frame. End with a perfectly symmetrical wide shot of both warriors holding swords above the…
- **Seedance 2.0** (video, by @AllaAisling) — 15-second cinematic animated short. A narrow wooden suspension bridge crosses a spectacular mountain gorge. A young woman walks across the bridge. Suddenly, one suspension cable snaps. The bridge swings violently sideways. The woman drops to her knees and grips the wooden ropes. Several loose boards fall away beneath her. CUT TO: A gigantic dragon flies upward through the gorge. The dragon rises directly underneath the bridge. It catches the underside of the bridge with both front claws, gripping the heavy structural beams beneath the wooden walkway. The dragon remains completely below the bri…
- **Nano Banana Pro** (image, by @Preda2005) — Create a 3D isometric diorama illustration of an original character named [CHARACTER NAME], peacefully working or practicing a creative activity in a nature-inspired environment, sitting at a wooden desk near a large open window with sunlight flowing in and soft greenery visible outside. The character should be inspired by [ARCHETYPE / PROFESSION – e.g. writer, painter, meditator, calligraphy artist, tea master] and have a unique visual identity, including soft, natural clothing, organic textures, and a calm, introspective personality. The entire scene must have a handcrafted felt aesthetic, w…
- **Nano Banana Pro** (image, by @Naiknelofar788) — Surrounding the realistic main subject are multiple cute, 3D-style chibi miniatures of the same person, with identical facial features, hairstyle, body proportions, and outfit. The chibi figures are naturally distributed around the subject, interacting playfully with her or nearby elements in a charming, non-intrusive way. Overlay the image with vibrant, hand-drawn doodle effects: soft white outlines around the subject, playful sparkles, doodle hearts, tiny flowers, smiley icons, and floating white handwritten phrases like "shine", "bright day", and "happy". The style seamlessly blends hyper-r…
- **Nano Banana Pro** (image, by @Arminn_Ai) — An isometric 3D cube-shaped miniature room (shallow cutaway true cube; everything strictly contained within the cube). The room is [Room Description]. Character: a chibi/figurine-style — [Character Description]. The character is [Action], with a [Expression] expression. Figure material looks like matte PVC, with big head / small body proportions. Lighting: [Atmosphere Name]: [Light Sources]; realistic reflections and colored shadows. Camera: slightly elevated isometric three-quarter view, front cube edge centered; no elements protruding outside the cube. Photoreal materials with fine detail; n…
- **Nano Banana Pro** (image, by @Gdgtify) — 2x2 grid, 16:9, do this for 4 famous cities: M₁ PERSONIFIED ICON MASCOT (master) the place's most iconic figure/monument personified as a friendly mascot character (face, hair, pose), central and dominant. ¬M₁ = no mascot / generic person / monument not personified / mascot not of the place. M₂ COSTUME = CITY COLLAGE the mascot's garment/body is a collage of the city's skyline/landmarks/districts (buildings, bridge, subway, park, road); the costume reads as the place's map. ¬M₂ = plain costume / collage unrelated / single landmark. M₃ ANTHROPO‑ICON FRIENDS ≥4 of the city's symbols (food, trans…
- **Nano Banana Pro** (image, by @harboriis) — A high quality mixed media portrait of a trendy young man (using my uploaded person photo) standing on a train station platform at night. The subject is posing coolly, holding the strap of an olive backpack with one hand. He is wearing a stylish streetwear outfit with an oversized layered black hoodie, gray sweatpants, and chunky sneakers. The background features a realistic KRL Jabodetabek Commuter Line train side view with its signature red striping, with the train doors wide open, bathed in bright cinematic station lighting and city night ambiance. Surrounding the realistic main subject are…
- **GPT Image 2** (image, by @Naiknelofar788) — Create a stylish modern travel poster inspired by risography about [City Name]. Use simplified illustrated images of the city: the Gwanganje Bridge, ocean waves, seagulls, beach umbrellas, fish baskets at the market, neon street signs and small abstract decorative figures. Create a bold, multi-layered risographic poster with imperfect ink design, translucent colors printed on top, graininess in semitones, rough paper texture, flat graphic shapes, playful geometric shapes, and energetic editorial composition. Use fluorescent pink, bright blue, warm yellow, orange-red, leafy green, and rich coba…
- **GPT Image 2** (image, by @SimplyAnnisa) — Create a stylish modern travel poster inspired by risography about [City Name]. Use simplified illustrated images of the city: the Gwanganje Bridge, ocean waves, seagulls, beach umbrellas, fish baskets at the market, neon street signs and small abstract decorative figures. Create a bold, multi-layered risographic poster with imperfect ink design, translucent colors printed on top, graininess in semitones, rough paper texture, flat graphic shapes, playful geometric shapes, and energetic editorial composition. Use fluorescent pink, bright blue, warm yellow, orange-red, leafy green, and rich coba…
- **GPT Image 2** (image, by @Taaruk_) — Create a vertical cozy home scrapbook poster inspired by the reference images. Combine a beautiful realistic interior photograph on the top half with a hand-drawn illustrated memory journal on the bottom half. Top section: A warm, bright and minimalist cozy home interior with natural cream walls, light wooden furniture, open wooden shelves filled with tiny ceramic houses, cute collectible figurines, miniature animals, books, framed artwork and handmade décor. Include a wooden desk or dining table, vintage radio, small lamp, flowers, baskets, ceramics and carefully arranged little objects. Soft…
- **GPT Image 2** (image, by @Naiknelofar788) — Create a sophisticated minimalist editorial art poster featuring [Subject] as the central subject, inspired by premium travel photography and refined hand-drawn illustration. Preserve the subject’s recognizable identity, proportions, silhouette, defining details, and visual character. Combine realistic elements with elegant artistic interpretation. Use a warm ivory paper background with a limited vintage color palette inspired by the subject’s location or personality. Add delicate ink outlines, architectural sketch lines, subtle hatching, halftone texture, faded print effects, and soft waterco…
- **GPT Image 2** (image, by @Goodmanprotocol) — Create a premium 4:5 vertical Instagram travel poster for [Country], designed as part of a cohesive collectible “Souvenir Box” series. Show an open vintage cardboard travel box containing exactly 10 culturally authentic souvenirs from [Country]: 1 hero object, 2–3 supporting objects, and additional smaller keepsakes. Choose objects with varied materials, sizes, shapes, and cultural significance. Include exactly 4 paper elements: - A folded illustrated map showing 4 different destinations - A perforated vintage travel ticket - A die-cut [Country] destination sticker - One culturally appropriate…
- **GPT Image 2** (image, by @user_ed334ac7) — Create a stylish illustrated digital collage / character mood-board of a fashionable young African man in a playful sketchbook aesthetic. The composition should feature multiple poses and expressions of the same man arranged across a clean white background like a personalized personality poster or lifestyle board. Include one full-body central pose as the main focus, surrounded by smaller poses, selfies, close-up portraits, sitting poses, relaxed candid moments, and a cute chibi version.He should wear:tinted fashionable sunglassesfitted neutral-toned t-shirt or sleeveless topslim dark jeanstre…
- **GPT Image 2** (image, by @user_1233d897) — A cutting-edge commercial 3D isometric illustration of a modern creative workspace, designed for a premium SaaS website hero section. Behance and Dribbble trend-setting style. The composition features a stylized, soft-edged hovering desk in the center. On the desk rests a clean, futuristic glass tablet glowing with a soft holographic interface, surrounded by floating geometric elements like 3D spheres, sleek metallic cylinders, and matte clay-textured coffee cups. High-end C4D rendering, Octane Render, claymation aesthetic with glossy and matte contrasts. Dopamine color palette featuring vibra…
- **GPT Image 2** (image, by @user_b9ad198a) — Create a highly detailed, premium travel-art illustration featuring a hand holding a white HỒ CHÍ MINH Transit Pass in the foreground. The ticket must look like a real transit ticket with visible text elements including: city name “HỒ CHÍ MINH”, ticket type “HỒ CHÍ MINH Transit Pass”, fare amount, validity date, and a realistic black QR code on the right side. The ticket acts as a portal to HỒ CHÍ MINH, with a realistic miniature city scene emerging from the ticket cutout. A signature transportation element — the city’s most recognizable local vehicle, train, tram, bus, or monorail — breaks ou…
- **GPT Image 2** (image, by @user_3a66c469) — PROMPT:A premium editorial infographic poster in 16:9 horizontal format about Ancient Rome, designed like a luxury illustrated history atlas page, with a warm ivory parchment background, deep navy serif typography, and subtle gold accent details, featuring a dramatic central floating miniature cityscape of ancient Rome with the Colosseum, Roman forums, marble temples, triumphal arches, roads, statues, and layered hills in a cinematic diorama composition with soft clouds and golden sunlight, highly detailed and realistic but stylized; on the left side a refined vertical information column with…

## Whisk AI 价格

## Subscriptions (monthly or yearly, cancel anytime)

- **Starter: $29.99/month or $288.00/year** — 2,400 credits per month.
- **Pro: $69.99/month or $624.00/year** — 5,200 credits per month.
- **Ultimate: $149.99/month or $1308.00/year** — 10,900 credits per month.
- **Crew: $79.00/month or $708.00/year per seat (min 3 seats)** — 5,900 credits per month.
- **Business: $119.00/month or $948.00/year per seat (min 3 seats)** — 7,900 credits per month.

## Credit packs (one-time, credits never expire)

- **Starter: $9.90** — 600 credits, never expire.
- **Creator: $29.90** — 2,200 credits, never expire.
- **Studio: $49.90** — 4,200 credits, never expire.
- **Business: $99.90** — 10,000 credits, never expire.

## Credit usage

| Model | Resolution | Credits |
| --- | --- | --- |
| Veo 3.1 Fast | 720p | 81 per video |
| Veo 3.1 Fast | 1080p | 87 per video |
| Veo 3.1 Fast | 4K | 241 per video |
| Veo 3.1 Quality | 720p | 335 per video |
| Veo 3.1 Quality | 1080p | 341 per video |
| Veo 3.1 Quality | 4K | 495 per video |
| Veo 3.1 Lite | 720p | 41 per video |
| Veo 3.1 Lite | 1080p | 47 per video |
| Veo 3.1 Lite | 4K | 201 per video |
| Seedance 2.5 | 480p | 38 per second |
| Seedance 2.5 | 720p | 85 per second |
| Seedance 2.5 | 1080p | 212 per second |
| Seedance 2.5 | 4K | 336 per second |
| Seedance 2.5 Flash | 480p | 25 per second |
| Seedance 2.5 Flash | 720p | 55 per second |
| Seedance 2.0 | 480p | 26 per second |
| Seedance 2.0 | 720p | 55 per second |
| Seedance 2.0 | 1080p | 137 per second |
| Seedance 2.0 | 4K | 279 per second |
| Seedance 2.0 Fast | 480p | 16 per second |
| Seedance 2.0 Fast | 720p | 34 per second |
| Seedance 2.0 Mini | 480p | 6 per second |
| Seedance 2.0 Mini | 720p | 11 per second |
| Kling 4.0 | 720p | 39 per second |
| Kling 4.0 | 1080p | 57 per second |
| Kling 4.0 | 4K | 133 per second |
| Kling 4.0 Flash | 720p | 23 per second |
| Grok Imagine | 480p | 4 per second |
| Grok Imagine | 720p | 7 per second |
| Grok Imagine | 1080p | 11 per second |
| Nano Banana Pro | 1K | 25 per image |
| Nano Banana Pro | 2K | 25 per image |
| Nano Banana Pro | 4K | 33 per image |
| Nano Banana 2.1 | 1K | 6 per image |
| Nano Banana 2.1 | 2K | 9 per image |
| Nano Banana 2.1 | 4K | 13 per image |
| Nano Banana 2 | 1K | 11 per image |
| Nano Banana 2 | 2K | 17 per image |
| Nano Banana 2 | 4K | 25 per image |
| Nano Banana 2 Lite | 1K | 6 per image |
| Nano Banana | 1K | 6 per image |
| GPT Image 2.5 Flare | 1K | 9 per image |
| GPT Image 2.5 Flare | 2K | 14 per image |
| GPT Image 2.5 Flare | 4K | 22 per image |
| GPT Image 2.5 Sunburst | 1K | 9 per image |
| GPT Image 2.5 Sunburst | 2K | 14 per image |
| GPT Image 2.5 Sunburst | 4K | 22 per image |
| GPT Image 2 | 1K | 9 per image |
| GPT Image 2 | 2K | 14 per image |
| GPT Image 2 | 4K | 22 per image |
| GPT Image 2.6 Economy | 1K | 7 per image |
| GPT Image 1.5 | 1K | 6 per image |
| GPT Image 1.5 | 2K | 30 per image |
| GPT Image 1 (4o) | 1K | 9 per image |

Credits are only charged for successful renders. Refunds: within 7 days of purchase if no more than 10% of the pack's credits were used.

## Whisk AI 常见问题

### 什么是 Whisk AI？

Whisk AI（whiskai.net 上的 WhiskAI）是一个在线图片混合工具。你最多提供三张图：主体、场景和风格，再加一句可选的提示词，它就会用 Nano Banana 或 GPT Image 生成一张融合三者的新图片，之后还能做成短视频。

### 这是 Google 的 Whisk 吗？

不是。WhiskAI 是一个独立服务，提供类似 Whisk 的主体、场景、风格混合流程。它与 Google 无关联，也未获其认可或赞助；Whisk 和 Google 是 Google LLC 的商标。

### Google Whisk 现在怎么样了？

Google 于 2024 年 12 月以 Google Labs 实验项目的形式推出了 Whisk。Labs 的工具经常调整，最新状态请以 labs.google 为准。如果你想要可自选 Nano Banana 或 GPT Image、统一积分的混合流程，Whisk AI 是一个独立的替代选择。

### 一定要写提示词吗？

不用。至少放一张图片就能点“生成”，Whisk AI 会告诉模型每张图的用途。写一句简短要求，比如“做成珐琅徽章”，能更好地控制风格。不放图时，只写提示词就是普通的文生图。

### 最多能混合几张图？

三个带标签的槽位分别对应主体、场景和风格。Nano Banana Pro 等模型总共支持最多 8 张参考图，需要时可以在槽位下方追加更多参考图。

### Whisk AI 用的是哪些模型？

图片混合使用 Google 的 Nano Banana 系列（默认 Nano Banana Pro）或 OpenAI 的 GPT Image 系列。“做成视频”使用 Seedance、Veo 3.1、Kling 4 或 Grok Imagine。每个结果都会标明实际使用的模型。

### Whisk AI 免费吗？

新账号赠送免费积分，足够完成第一次混合。之后按张使用积分，可以买积分包或订阅；点击“生成”前按钮上会显示准确的消耗，失败会自动退还积分。

### Whisk AI 混合的图可以商用吗？

可以。购买任意积分包或 Pro 及以上套餐，按照服务条款即可商用。请确认你对上传的照片拥有权利，尤其是他人肖像、商标以及不属于你的角色。

### 能把混合图做成视频吗？

可以。在任意生成好的图片上点“做成视频”，视频工作室会把它设为首帧；描述动作并选择 Seedance、Veo 3.1、Kling 4 或 Grok Imagine 即可。

### 我上传的图片是私密的吗？

上传的图片会被保存以便模型读取，只用于你自己的生成。生成结果默认会出现在公开画廊；付费套餐提供私密模式，可只自己可见。


---

Source: https://whiskai.net/zh · Site index for AI agents: https://whiskai.net/llms.txt · Full content: https://whiskai.net/llms-full.txt
