Generate images and edit photos with Nano Banana 2, Google’s Gemini 3.1 Flash Image model. Explore character sheets, product visuals, or a new style using a written brief and optional references.
Explore the tasks shown by these actual Nano Banana 2 results. Copy a prompt to try in the generator, or open the linked case for its saved setup. Supply your own images or clips when the request uses references.
Reference Character Views and Expressions
Build a character sheet from a person reference, arranging different poses and expressions within one image.
Please create a detailed character design sheet based on the person in this photograph.
As this will be used for a live-action film, please ensure it is not produced as stylized artwork.
It must include the following three character elements. Please delete the original image of the person upon completion.
Character multi-angle (three-view) turnaround sheet
Expression sheet showing basic emotional states: joy, sadness, anger, surprise, fear and neutral
Pose sheet showing typical actions: e.g., the character running, jumping, laughing and jumping
*Note: The word "jumping" appears twice in the original text, which may be a typo. If it is intended to refer to different actions, common alternatives include "walking", "sitting" or "gesturing".
Create a fashion composition around a person and electric scooter from a written scene brief.
{
"image_settings": {
"aspect_ratio": "3:4",
"resolution": "768/1024"
},
"prompt": {
"scene_and_composition": {
"visual_hook": "A stylized urban youth leaning casually against a sleek electric scooter, emphasizing a relaxed 'street-style' silhouette against a high-contrast white backdrop.",
"camera_angle": "Slight low-angle 'hero shot' to accentuate the scooter and the subject's stance.",
"framing_and_depth": "Full-body shot, centered composition, high-key studio lighting with absolute white background, sharp focus throughout.",
"medium_physics": "Professional product photography style, high-contrast, sharp digital clarity, clean edge definition, no grain."
},
"subject_content": {
"identity_features": "Young male, mixed ethnicity, braided dreadlock hair texture, angular facial features, sharp jawline, tan skin tone, athletic build.",
"overall_pose": "Standing one-footed on a black electric scooter, one hand on the handlebar, other hand in cargo pocket, slight lean, cool effortless posture.",
"limb_articulation": "Right arm extended down gripping the handle, left arm bent at elbow with hand tucked into cargo short pocket, right leg balanced on scooter deck, left leg hovering slightly forward for balance.",
"pose_dynamics": "Center of gravity shifted toward the rear of the scooter, casual lean, relaxed muscle tension.",
"makeup_look": "Natural skin texture, clean-cut, post-grooming aesthetic, no visible makeup.",
"emotional_vibe": "Confident, nonchalant, urban-cool, detached, trend-focused.",
"main_object/apparel": "Oversized mustard-orange graphic t-shirt with screen-printed art; baggy, cream-colored technical cargo shorts with visible pocket flaps and branding patches; white ribbed crew socks; chunky 'dad-style' designer sneakers.",
"hair_and_styling": "Braided dreadlocks held in place, white over-ear wireless headphones resting around the neck/ears, layered silver chain necklaces.",
"interaction_with_environment": "Weight of the subject compressing the deck of the electric scooter, feet firmly planted/posed on the grip tape, shadow cast directly beneath the wheels."
},
"environment": {
"visual_depth_stack": "Foreground: 1 black electric scooter with orange rim accents; Midground: 1 male subject in orange shirt and cargo shorts, 1 pair of white chunky sneakers, 2 white ribbed socks; Background: 1 pure white infinite studio backdrop.",
"lighting_atmosphere": "Key: bright, diffuse studio light from top-front; Ambient: high-key fill, near-zero shadows on background; Interaction: sharp shadow cast on floor from scooter wheels, high specular highlights on the metal and plastic surfaces of the scooter."
},
"realism_rules": [
"Catalog-style high-key studio photography",
"Perfectly white, seamless background",
"Focus on texture of fabric and plastic",
"Sharp edge outlines",
"Authentic urban streetwear proportions",
"No blurred or depth-of-field effects on the subject"
]
},
"negative_prompt": [
"3D render",
"cartoon",
"painting",
"blurry background",
"vignette",
"low-resolution",
"grainy",
"flat lighting",
"distorted anatomy",
"floating subject",
"amateur composition",
"cluttered environment",
"text artifacts",
"oversaturated colors"
]
}
Build a styled product photograph around the uploaded perfume, specifying the bottle details and new presentation.
Use the uploaded perfume as the exact product reference. Preserve bottle shape, smoked-plum glass, black cap and every label word. Build a premium 4:5 Halloween beauty advertisement: the single bottle stands on a dark oxblood stone pedestal, surrounded by one sculptural black pumpkin, thin curling smoke and a few wax-dripped candles kept behind the product. Warm amber rim light and cool plum shadows, polished commercial realism, label fully readable. Add only one headline in the upper third, spelled exactly: ‘MIDNIGHT BLOOMS’. No extra product, no label redesign, no bats, no clutter, no watermark.
Nano Banana 2 is Google’s Gemini 3.1 Flash Image model, positioned for efficient image generation and conversational editing.
Efficient Image Creation and Editing
Google positions this Flash model for efficient, high-volume image work. Both new-image briefs and reference-led changes belong to the model’s image workflow.
Words and Pictures in a Designed Layout
The model supports graphics that combine text with imagery. Posters and character sheets need a specified composition as well as a description of the subject.
More Room for Tall and Wide Ideas
Google documents additional tall and panoramic aspect ratios. WeShop’s listed formats determine which canvas you can choose on this page.
WORKFLOWS
Choose Your Starting Input
Use the Nano Banana 2 input mode that matches the material you have.
Start from Your Reference Photos
Upload the subject or scene for the edit. For several inputs, identify what each reference supplies and describe the requested change.
Create from a Written Brief
Describe a new subject, setting, composition, and visual treatment. Include any words that belong in the image.
PROMPT GUIDE
How to Direct Nano Banana 2 with a Prompt
Learn from the actual instructions behind these results. Each example keeps its complete original prompt and available inputs. For a new request, replace its subject and reference assignments with your own.
Describe the Marks or Medium
Name the visible treatment, such as the drawing marks or material, rather than relying only on a mood word. In this case: “Convert this image into a delicate watercolor painting with soft flowing edges, pastel hues, visible paper texture, and loose artistic brushwork.”
Convert this image into a delicate watercolor painting with soft flowing edges, pastel hues, visible paper texture, and loose artistic brushwork.
Describe a Visible Material
Use concrete surface or material details to define the requested look. In this case: “Hard direct flash, sharply defined hand shadow, glossy pale-pink manicure, realistic skin texture, premium social-commerce styling, 3:4 portrait.”
Create a vertical commercial beauty photograph of a hand holding three fictional, unbranded peptide lip-tint tubes in blush pink, cocoa brown and dusty rose against a warm off-white background. Add scattered Y2K acrylic rhinestones to the tube bodies: pink hearts, clear teardrops, purple flowers, silver stars, tiny round crystals and a few pearls. Gems naturally conform to curved surfaces with transparent facets, flash highlights, tiny contact shadows and subtle adhesive traces. Keep caps, tube geometry, color blocks and minimal non-readable label structure crisp. Hard direct flash, sharply defined hand shadow, glossy pale-pink manicure, realistic skin texture, premium social-commerce styling, 3:4 portrait. No real logos, no watermark, no unrelated or legible text.
Place the Light and Color in the Scene
Specify the visible lighting or color arrangement as part of the composition. In this case: “The lighting highlights their figures, giving the scene a dramatic and cinematic feel.”
This image shows a man sitting at a desk in a dimly lit office, speaking on a corded telephone. He is wearing a white dress shirt and a dark red tie, and appears serious and focused as he looks down at the papers in front of him.
The desk is covered with stacks of documents, folders, and files, suggesting a busy or bureaucratic work environment. Behind the man stands a large white horse, positioned calmly inside the office. The horse’s presence is unexpected and surreal, creating a strong contrast with the formal setting.
The background is dark and minimal, which draws attention to the man, the desk, and the horse. The lighting highlights their figures, giving the scene a dramatic and cinematic feel.
Overall, the image presents a surreal and symbolic composition, combining ordinary office work with an unusual element. It suggests themes of absurdity, mystery, or quiet tension, as if something extraordinary is hidden within an otherwise serious workplace setting.
HOW TO USE
Create Your First Nano Banana 2 Image
Start in the generator on this page. To reuse an example, click Copy Prompt, return to the description field, and adapt the prompt to your own inputs.
Provide the Scene and Inputs
Write your brief and add the reference media required by the selected mode. Explain the role of each supplied input.
Choose the Output Options
Choose the aspect ratio, image size, quality, and image count where offered. Check the form’s displayed points before submitting.
Generate and Review
Click Generate and review the resulting image. Check composition, required words, and the subject details that matter to your task. Adjust your prompt or references for the next attempt.
Compare the documented model roles and the inputs and options currently listed in WeShop. Model names or matching settings do not by themselves establish equal output quality or a fixed generation-time difference.
Dimension
Nano Banana
Nano Banana Pro
Nano Banana 2
Nano Banana 2.1
Model role
Original Gemini 2.5 Flash Image creation and editing
Gemini 3 Pro Image for detailed design and creative control
Gemini 3.1 Flash Image for efficient creation and editing
Gemini 3.6 Flash update for layouts and visual continuity
Inputs listed for the page
Start from Your Reference Photos; Create from a Written Brief
Create a New Design; Use a Reference Subject
Start from Your Reference Photos; Create from a Written Brief
Start from Your Reference Photos; Create from a Written Brief
Consider Nano Banana 2 for frequent image creation and editing. Compare Pro for demanding design control, or 2.1 for its later layout and continuity updates.
MORE IMAGE MODELS
Explore More Image Models
Open a related model page to review its capabilities, examples, and current generation options.
Answers about this version, its inputs, and how to get started.
What is Nano Banana 2?
Nano Banana 2 is Google’s Gemini 3.1 Flash Image model, positioned for efficient image generation and conversational editing.
What is a good task to consider it for?
Consider Nano Banana 2 for frequent image creation and editing. Compare Pro for demanding design control, or 2.1 for its later layout and continuity updates.
What do I need to provide?
Start from Your Reference Photos. Upload the subject or scene for the edit. For several inputs, identify what each reference supplies and describe the requested change. Create from a Written Brief. Describe a new subject, setting, composition, and visual treatment. Include any words that belong in the image.
Which output settings are offered here?
Image size: 1K, 2K, 4K
How do I try an example prompt?
Click Copy Prompt, return to the generator, and paste it into the description field. Add your own required references and update their roles. Linked case previews and Use This Prompt open the saved case; they do not automatically fill this page’s form.
What should I know about this version?
Google’s vendor tools do not all appear in this form. Search and Thinking controls are not implied by the model name.
