Complex image instructions
Describe subjects, composition, lighting, materials, labels, and constraints in one structured brief.
Google image model · Live workspace
Nano Banana Pro image generator
Generate or edit detailed images with Google’s Gemini 3 Pro Image model. Use text prompts, upload visual references, and choose the output settings currently supported by this platform.
01
Create with Nano Banana Pro
This workspace keeps Nano Banana Pro selected so you can test the model without switching away from this page.
Results
Generated content will appear here
02 · Overview
Verified model and platform facts
Facts reviewed:
- Provider
- Official model ID
- gemini-3-pro-image
- Inputs
- Text and images
- Platform output sizes
- 1K, 2K, 4K
- Platform reference limit
- Up to 14 images
- Platform status
- Live
What is Nano Banana Pro?
Nano Banana Pro is Google’s product name for Gemini 3 Pro Image. Google positions it for professional asset production and complex instructions, with image generation and editing, thinking, and Google Search grounding listed in its model documentation.
On this website, Nano Banana Pro is available through a focused online image generator and editor. Google defines the underlying model capabilities; this platform separately defines the controls exposed here, including supported aspect ratios, reference-upload limits, output sizes, and point estimates.
Choose this model when your brief has several visual constraints, readable in-image text, detailed layouts, product mockups, or fact-aware graphics. Generated details can still be incorrect, so names, figures, logos, labels, and factual diagrams should be checked before publication.
03 · Capabilities
Where Nano Banana Pro is useful
Capabilities below reflect Google’s documentation; available controls refer specifically to this platform.
Image generation and editing
Start from text alone or upload images to guide a new composition or targeted visual change.
Text-heavy visual concepts
Explore posters, menus, diagrams, and campaign layouts where legible text placement matters.
Multiple output sizes
The current workspace exposes 1K, 2K, and 4K choices; point estimates update before submission.
Flexible canvas formats
Choose square, portrait, landscape, and ultrawide ratios supported by the current integration.
Reference-led direction
Upload references for products, characters, palettes, or composition while reviewing each output for drift.
Character and brand consistency
Use multiple references to guide recurring characters, products, and visual systems across a set of assets.
Search-grounded visual research
Google documents Search grounding as a model capability for fact-aware visuals; its availability can depend on the integration and task.
04 · Use cases
Common use cases
Capabilities below reflect Google’s documentation; available controls refer specifically to this platform.
01
Campaign key visuals
Explore polished hero concepts with defined art direction, copy space, and channel-specific framing.
02
Product mockups
Place a product reference into new environments while checking shape, labels, and packaging carefully.
03
Infographics and diagrams
Prototype information-rich visuals, then verify every fact and replace critical text in a design tool if needed.
04
Localized creative variants
Draft region-specific layouts and in-image language variants without assuming text will be error-free.
05
Social and editorial graphics
Create thumbnail, cover, presentation, and article concepts in the aspect ratio required by each channel.
06
Character-led storyboards
Use reference images and explicit continuity notes to explore recurring characters across multiple scenes.
05 · Prompts
Nano Banana Pro prompt examples
Use concrete constraints and treat each result as a draft that still needs review.
Prompt 01
Product launch visual
Control material, lighting, layout, and copy space.
Create a premium studio product photograph of a matte green travel bottle on pale limestone. Soft morning side light, realistic condensation, restrained shadows, wide negative space on the left for campaign copy, 4:5 composition, no visible brand logo.
Prompt 02
Educational infographic
Request a clear hierarchy and limited factual scope.
Design a clean vertical infographic explaining the four stages of a butterfly life cycle. Use four clearly separated panels, concise English labels, natural-history illustration, cream background, dark green headings, and accessible contrast.
Prompt 03
Reference-based redesign
Upload a source image before trying this editing brief.
Keep the uploaded product shape and camera angle unchanged. Replace the background with a quiet winter window scene, add soft blue-hour reflections, preserve label readability, and leave all product proportions intact.
Prompt 04
Multilingual campaign system
One visual direction adapted into coordinated campaign formats.
Create a coordinated launch campaign for a premium skincare serum: 4:5 social visual, 16:9 web banner, and square retail tile. Keep the same bottle, ivory-and-amber palette, soft botanical shadow, and art direction. Reserve distinct copy areas for English and Traditional Chinese; do not render long paragraphs.
Prompt 05
Editorial location diagram
A polished information-led visual with spatial relationships.
Design an elegant illustrated visitor map for a compact waterfront arts district. Show five numbered venues, walking connections, ferry pier, green space, and a simple legend. Warm editorial palette, refined linework, clear hierarchy, landscape layout, no invented street names.
06 · Workflow
How to use Nano Banana Pro
- 01
Describe the outcome
State the subject, scene, composition, lighting, visual style, required text, and anything that must not change.
- 02
Add references if needed
Upload clear source images only when they materially guide identity, product shape, palette, or layout.
- 03
Choose platform settings
Select an aspect ratio and output size, then review the displayed point estimate before submitting.
- 04
Review and refine
Check text, anatomy, factual content, product details, and brand elements before iterating or downloading.
Responsible use
Limits and important distinctions
Model documentation describes what Gemini 3 Pro Image can support in Google’s APIs. Google documents up to 14 reference images for this model family, with separate object, character, and style-reference limits. This platform currently accepts up to 14 references, but its controls, availability, queue time, and point costs are independent and may change.
Image models may invent text, objects, citations, logos, or visual facts. Reference images improve direction but do not guarantee exact identity or geometry. Use human review and obtain appropriate rights for uploaded material and published outputs.
Google states that generated images include a SynthID watermark. Search grounding and the model’s internal thinking process can improve planning or factual context, but neither guarantees factual accuracy or a production-ready result.
Sources
Sources and review boundary
Provider facts were checked against Google’s current documentation. Platform settings come from the live workspace configuration.
07 · FAQ
Nano Banana Pro FAQ
What is the official model behind Nano Banana Pro?
Google identifies Nano Banana Pro as Gemini 3 Pro Image, with the model ID gemini-3-pro-image.
Can Nano Banana Pro edit an uploaded image?
Yes. Google documents text and image inputs for generation and editing, and this workspace currently accepts reference uploads.
Does this page use Nano Banana Pro directly?
The workspace selects the platform model option named Nano Banana Pro by default. You can also switch to another available image model from the model menu.
Can I generate 4K images?
The current platform integration exposes 1K, 2K, and 4K output choices. Availability and point estimates are shown before submission.
Is Nano Banana Pro guaranteed to render text correctly?
No. It is designed for stronger text rendering, but spelling, layout, and factual accuracy still require review.
What is the difference between Nano Banana and Nano Banana Pro?
Nano Banana is also used as Google’s broader name for Gemini native image generation. Nano Banana Pro specifically refers to the higher-capability Gemini 3 Pro Image model for complex instructions and professional asset workflows. Available speed, price, and controls depend on the platform.
How many reference images can I use?
Google documents up to 14 references for Gemini 3 Pro Image, divided across object, character, and style categories. This workspace currently accepts up to 14 images, although platform limits can change.
How should I write a Nano Banana Pro prompt?
State the subject, composition, camera or viewpoint, lighting, materials, required text, aspect ratio, references, and anything that must remain unchanged. Put critical constraints ahead of decorative style terms.
Is Nano Banana Pro free to use here?
This platform uses points rather than promising unlimited free generation. The current estimate appears in the workspace before you submit a task and can change with the selected settings.
Is this an official Google website?
No. Nano Banana Pro is an independent multi-model platform. Google owns the Gemini and Nano Banana model names.
Related models and tools
01
Nano Banana 2
Compare Google’s general-purpose Flash image model on its dedicated model page.
02GPT Image 2
Open the verified GPT Image 2 generation and editing page.
03AI image generator
Compare the image models currently available on the platform.
04Image to image
Start from an existing image and guide the transformation.
Create next
Try Nano Banana ProCreate with Nano Banana Pro
Return to the workspace with Nano Banana Pro selected by default, choose the settings you need, and review the result before using it.