Nano Banana AI 2: Revolutionary AI
Nano Banana AI 2 offers 50 practical prompts for the new image model based on Google's Gemini family. It enables the creation of product shots, social media graphics, architectural renderings, and infographics.
Introduction & Basics
Nano Banana AI 2 is an image model from Google's Gemini family. It includes Nano Banana (based on Gemini 2.5 Flash Image) ) and Nano Banana Pro (based on Gemini 3 Pro Image). ). Platforms like BananaNano.ai, Atlabs and Pixlr integrate these models into their editors, emphasizing their spatial understanding, text fidelity, and consistent editing of image series.
Google describes Nano Banana as an image model that combines text and image prompts to generate detailed, realistic, or playful subjects. It supports local edits, multiple reference images, and conversational mode via the Gemini-App and the Google AI Studio.
The Pro variant, based on Gemini 3 Pro Image, , offers improved text rendering, search-based world knowledge, finer camera and lighting control, and the ability to mix up to 14 reference images and up to five people in one subject ( (Google).
Independent platforms like BananaNano.ai summarize Nano Banana 2 and Nano Banana Pro as Google models with strong visual reasoning, 3D object manipulation, improved text fidelity, and context-aware image editing. Tool directories like GptDemo.net and Tools-AI Online list 'Nano Banana AI-2' as an AI image editor for text-to-image, image-to-image, background generation, and photo enhancements.
Pixlr integrates Nano Banana as a chat-based image editor where users can upload photos and describe desired changes in everyday language. Reports from The Verge, Android Central and TechRadar highlight that Nano Banana Pro is directly available in the Gemini app and is receiving integrations in AI Mode in Search, NotebookLM, and the Gemini API.
Areas of Application
BananaNano.ai describes use cases for Nano Banana 2 ranging from product design and marketing to social media content, digital illustration, and photo restoration, with a focus on precise control over lighting, perspective, 3D objects, and consistent image series.
Pixlr shows that Nano Banana is suitable for recurring image patterns, such as selfies with fictional stars, 3D figurines, LEGO style, Ghibli looks, object removal, and classic retouching.
TechRadar and Android Central emphasize the strengths of Nano Banana Pro for text-heavy images like infographics, diagrams, and posters, as fonts become more legible in multiple languages and contrast, mood lighting, and camera angles can be precisely controlled.
Prompt guides from Atlabs, Analytics Vidhya and Higgsfield utilize Nano Banana for tasks like blog-to-image, photo editing, infographics, logo design, and image blending, often with reference images and multi-part instructions for consistent series.
The GitHub-Sammlung „awesome-nano-banana“ and the Prompt-Bibliothek nanobananaprompt.org bundle community examples by style, subject, and technique.
Video tutorials on YouTube, such as „Nano Banana Pro: Your new creative partner“ or „Nano Banana AI Tutorial | How to Use Google's Free Tool“, , guide through the operation and demonstrate free access, 4K workflows, and creative series prompts.
Source: YouTube
Source: YouTube
Prompting Strategies
Google recommends clear subject descriptions, precise specifications for style, camera, lighting, perspective, and reference images for Nano Banana Pro to correctly interpret details and context.
Atlabs shows how up to 14 reference images can be used specifically to create consistent series.
Tom’s Guide tests Nano Banana Pro with trends like Funko Pop figures and aging effects to demonstrate the model's controllability.
The following prompts are based on real-life scenarios and can be used or adapted directly in Nano Banana AI 2.
Branding & Business
These prompts support the creation of visual content for marketing and business applications.

Source: fluxproweb.com
Interacting with Nano Banana AI: A Futuristic Banana at the Center of Digital Creation.
Generate a 4K product photo of a matte, black stainless steel water bottle on a light gray background, with soft studio lighting from the left, subtle reflections, and shadows, suitable for an Amazon product listing hero.
Design a key visual for a sustainable D2C brand: A reusable water bottle stands on a wooden pedestal, with blurred plants in backlight behind it, a clear, easily readable tagline text box in the upper right: 'Refill, not landfill.'
Create a LinkedIn cover image in landscape format 1584x396 pixels for a small consulting agency: minimalist isometric illustration set of three people at a large whiteboard with diagrams, blue tones, ample negative space on the left for text.
Design a series of three connected ad visuals for a SaaS product: each with the same UI layout on a laptop, tablet, and smartphone, consistent colors and typography, realistic office setting in the background.
Generate a product mockup for new coffee packaging: a matte stand-up pouch with a clearly visible logo in the center, next to it a cup of coffee with crema, warm morning light, wooden surface, and a slightly blurred kitchen background.
Take the uploaded photo of a business card, align it perspectively correct, remove visible fingers, replace the existing text with a clearly legible inscription 'Studio Nordlicht – Visual Design' in a modern sans-serif font.
Create a one-page A3 portrait poster for a local craft business: photo of the team in work clothes in front of their workshop, with a bold headline above, three keyword icons for services below, friendly, warm color mood.
Create a set of four square icons for a website navigation bar (Home, Services, References, Contact) in the same style: flat vector illustrations, dark background, clear outlines, slight glow around the icons.
Content & Social Media
These prompts help create engaging content for social media and other digital platforms.
Generate a vertical 9:16 cover image for an Instagram Reel about productivity: person at a desk with a laptop, dramatic side lighting, bold colors, large, easily readable text at the top: '5 Minutes, 1 Focus' in modern typography.
Design a 16:9 YouTube thumbnail for a video about Nano Banana AI 2: portrait of the presenter in studio look on the left, abstract collage of AI-generated images on the right, separated by a yellow banner with the inscription 'AI Image Power'.
Take the uploaded screenshot of a blog article and create a social media share image from it: shortened headline, key takeaway quote extracted from the text in large font, background as a blurred version of the screenshot.
Generate a 3-image Instagram carousel set for a data storytelling article: Slide 1 with a concise headline, Slide 2 with a simplified diagram, Slide 3 with a call-to-action 'Read more on the blog' and a click arrow icon.
Create a 3-image set for Pinterest about a recipe: photogenically staged plate on a rustic wooden table, close-up of the main ingredients, and a graphical overview with step-by-step icons, all in a consistent pastel color scheme.
Generate a 1200x900 pixel LinkedIn document cover for a whitepaper on AI introduction in the SME sector: abstract factory hall with schematic data flows, serious blue and gray tones, title in the upper third with plenty of white space.
Take an uploaded portrait photo and create a title graphic for a newsletter from it: crop to bust shot, transform background into a soft gradient design, circular cutout of the face, with clear title text and date next to it.
Generate a 14-image storyboard for a short commercial: same main character, changing scenes – office, street cafe, home office – consistent clothing and appearance, clear captions under each frame for later film editing.
People & Portraits
These prompts focus on editing and creating portraits and human representations.

Source: felloai.com
Creative Transformation: Nano Banana AI Transforms Portraits into New Scenarios.
Take an uploaded selfie and generate a series of four portraits of the same person in different lighting moods: golden hour, overcast daylight room, neon lights in the city, and soft studio beauty light, each with a neutral background.
Retouch the uploaded portrait subtly: keep skin texture realistic, remove only rough blemishes, sharpen eyes slightly, transform background into a soft bokeh, filmic color look but natural.
Transform the uploaded portrait into an analog-looking black and white photo: strong contrasts, slight film grain, darken background, lighting as in a classic 1950s studio portrait.
Create a series of three portraits of the same fictional person: once as a modern manager in an office, once casual in a cafe, once as a speaker on a stage, each with similar facial anatomy and recognizable hairstyle.
Take a full-body standing photo, replace the background with an urban rooftop terrace at sunset, adjust the person's shadows and reflections to the new lighting, add a blurred city skyline in the background.
Generate a digital oil painting version from an uploaded photo: rough brushstrokes, warm color palette, background abstracted into color fields, lighting as in a classic Rembrandt portrait.
Create a team photo from four individually uploaded portraits: people arranged side-by-side in a bright office, same lighting and color look for everyone, subtle company name on a glass wall in the background.
Transform a portrait into a keynote title graphic: person in the left half of the image, right half with a dark gradient background, large title text, and placeholders for date and location.
Products & E-Commerce
These prompts are designed for creating product images and visual content for online retail.
Generate three variants of the same sneaker in white, black, and beige on an identical background: softly illuminated studio, slight shadow towards the back, clean white floor, camera slightly from above at a 45-degree perspective.
Take an uploaded product photo of headphones and place it in a lifestyle scene: wooden desk with a laptop, houseplant, warm window light, depth of field so that the headphones remain clearly in focus.
Generate an Amazon product image set: main image with the isolated product on a pure white background, detail shot of buttons and ports, and an image with schematically drawn dimensions and labeled features.
Design a graphical comparison of two product variants: both devices side-by-side, with clearly structured text blocks below them with icons for battery, weight, size, and price range.
Generate a 3D mockup for a software box in shelf style: oblique perspective, reflective floor, background slightly blurred, title and subtitle easily readable, logo in the top left.
Take a flat product photo of a book cover and simulate a realistic 3D representation from it: bound book slightly open, shadow cast on a wooden surface, soft side lighting.
Create a TikTok Shop thumbnail for a beauty product: product in the foreground, blurred bathroom surface in the background, strong color contrast, round badge with the inscription '-20% today' in the corner.
Architecture & Spaces
These prompts enable the visualization and editing of architectural and space representations.

Source: user-added
A yellow background with three photos showing a transformation of two people into a couple by the pool.
Take a photo of an empty living room and create two split-screen variants: left the current state, right a modern Scandinavian interior design with lots of wood, light fabrics, and a large houseplant.
Generate aperspectival 3D visualization from an uploaded 2D floor plan sketch: bright, realistic living space with appropriately placed furniture, eye-level viewpoint, daylight from large windows.
Transform an AirBnB photo of a bedroom into a high-quality listing shot: correct lens distortions, smooth out the duvet, remove cables, add warm, uniform lighting and small decorative elements (books, plants).
Create an exterior visualization of an office building at evening light: glass facade reflects a city, some offices lit inside, light rain on the street, lens flare from streetlights.
Take a street view with parked cars and transform it into a car-free pedestrian zone with trees, benches, and people sitting or walking leisurely, lighting like on a warm spring evening.
Create three variants of the same interior: once as a modern office, once as a co-working space, and once as a creative studio, each with different furniture but the same room geometry and camera perspective.
Technology, Infographics & Documentation
These prompts are suitable for creating technical diagrams, infographics, and visual documentation.
Create a clear process diagram from a textual description of a business process: horizontal swimlanes for departments, numbered steps, subtle colors, plenty of white space, easily readable labels.
Take a photo of a whiteboard full of handwritten notes and sketches and transform it into a clean, digital infographic with clear arrows, icons, and clear typography.
Design a technical block diagram for a web application: browser, API, database, and external services as clearly separated blocks, connecting arrows, short, precise labels, color coding per layer.
Generate an illustrated guide in the style of an IKEA manual for assembling a simple shelf: four step-by-step images in black and white, figurative pictograms without text, clear depiction of screws and parts.
Take a photo of a physical circuit diagram and reconstruct a clean, digital version with clearly readable symbols, straight lines, and labeled components, neutral color palette.
Photo Editing, Compositing & Series
These prompts support advanced image editing, compositing, and the creation of consistent image series.
Remove distracting people in the background of the uploaded vacation photo, realistically reconstruct the sandy beach and the sea, adjust shadows and wave direction so that no editing traces remain visible.
Transform a summer landscape photo into a winter version: trees with snow, frozen lake, cold, blue light, visible breath clouds on people in the picture, sky slightly cloudy.
Extend an uploaded landscape format shot to the left and right to a wide panorama, maintain subject and perspective, organically add matching surroundings so that the image edge appears natural.
Take a night shot of a city street and add subtle motion blur effects to the car lights, increase contrast, make reflections visible on the wet street, cinematic look like in a neo-noir film.
Generate a series of four variants from an input image with the same person and pose, but different color looks: neutral, warm orange-teal, cool cine LUT, and retro film with visible grain.
Experiments & Fun
These prompts offer opportunities for creative and playful image experiments.
Transform an uploaded full-body photo of a person into a stylized 3D figure in the style of a collectible: large head, small body, glossy surface, person standing in an illuminated display case with soft light.
Generate a 16-bit pixel art version of an uploaded city photo: reduced color palette, clear edges, visible pixels, yet recognizable buildings and lights, suitable as a background graphic for a retro game.
Take a pet photo and transform the animal into a LEGO minifigure version: blocky shapes, typical LEGO facial features, small scene on a pedestal, soft studio lighting, neutral background.
Conclusion & Outlook
Nano Banana AI 2 combines Google's Gemini image models with editor interfaces like the Gemini-App, BananaNano, Pixlr and other integrations. This enables studio quality, series consistency, text fidelity, and conversational editing ( (Google, Higgsfield).
Success lies in the precision of the prompts. Google advises to concretely formulate subject, target medium, style, camera parameters, lighting, text content, and reference images, and to conduct image development as a dialogue.
For further steps, the official Prompting-Artikel von Google, der Atlabs-Guide, of the Analytics Vidhya practical examples from Higgsfield and „awesome-nano-banana“ , as well as community resources like nanobananaprompt.org are recommended. These offer examples, fine-tuning ideas, and workflows.
With clear, context-rich prompts like the 50 examples, Nano Banana AI 2 can be used for repeatable, consistent image series in projects, from social ads to product campaigns.