Nano Banana 3 Docs

Nano Banana 3 is an online AI image workspace for fast visual creation and controlled editing. It helps you turn prompts and reference images into polished visuals for ads, product scenes, posters, mockups, presentations, diagrams, and social content.

Use this guide when you want to understand the product, choose the right workflow, and get cleaner results from the homepage image generator.

What is Nano Banana 3?

Nano Banana 3 combines text-to-image generation, image-to-image editing, and reference-guided design in one browser workflow. The product is shaped around Nano Banana Pro-style strengths: clearer typography, richer real-world context, multi-image composition, localized visual ideas, and precise natural-language edits.

Instead of treating generation and editing as separate steps, Nano Banana 3 keeps the loop simple:

• describe the image you want
• upload references when identity, product shape, style, or layout should stay consistent
• choose a model, aspect ratio, quality, and output format
• generate several directions
• refine the best result with focused follow-up instructions
• review and export the final image

Core workflows

Text to image

Use text to image when you need a new visual from a prompt. It works best when the brief includes the subject, setting, style, camera angle, lighting, composition, required text, aspect ratio, and final use.

Good text-to-image tasks include:

• ad concepts and campaign visuals
• poster layouts and thumbnails
• product scenes and ecommerce graphics
• social posts in square, vertical, or wide formats
• diagrams, explainers, and education visuals
• editorial mockups and presentation images

Image to image

Use image to image when you already have a source image and want controlled changes. Upload the image, then explain what should change and what must remain recognizable.

This workflow is useful for:

• replacing or extending backgrounds
• restyling an existing scene
• adapting one image to another aspect ratio
• changing atmosphere, lighting, or finish
• creating variations while preserving a face, product, outfit, color palette, or layout

Reference-guided editing

Upload one or more references when consistency matters more than surprise. References can guide character identity, product resemblance, brand mood, material texture, scene structure, pose, or style direction.

The clearest prompts separate preservation from change:

• what to keep: face, logo placement, product shape, outfit, pose, color, composition
• what to change: background, camera crop, lighting, text, language, mood, season, format
• what to avoid: distorted hands, wrong wording, extra logos, changed product details, unreadable labels

How to use Nano Banana 3

The homepage workflow is designed around four practical steps.

1. Describe or upload

Start with a prompt, a source image, or reference images. If you need a fresh concept, write a complete prompt. If you need controlled editing, upload the image first and describe the target change.

For stronger results, include:

• subject and purpose
• scene or background
• camera angle and framing
• lighting and mood
• style or material direction
• exact in-image text, if needed
• final channel, such as ad, poster, product page, slide, or social post

2. Pick your output

Choose the mode, model, aspect ratio, quality, and output format before generating. This helps the result match the real destination instead of forcing you to fix the layout later.

Use square formats for feed posts and product cards, vertical formats for stories and posters, wide formats for banners and presentations, and input matching when the source image dimensions should stay close.

3. Refine with context

Nano Banana 3 performs better when your prompt has context. Add details about the audience, region, language, product use, design style, camera, typography, and composition.

For localized visuals, describe the language, setting, cultural cues, and market context you want. For explainers or diagrams, describe the structure and label placement instead of only naming the topic.

4. Review and export

Generate multiple versions, compare them, and refine the strongest one. Treat every AI image as a draft until it passes a final check.

Before publishing or handing off an image, review:

• text accuracy and spelling
• brand fit and logo behavior
• product or character consistency
• factual claims in diagrams or explainers
• likeness rights and usage permissions
• platform rules, content credentials, or watermark requirements

Key features

Readable text rendering

Nano Banana 3 is designed for visuals where text matters: posters, packaging, labels, signs, social graphics, thumbnails, diagrams, and localized layouts. Keep text short, quote exact wording, and describe placement and type style.

If a result is close but the words are wrong, run a focused edit that only corrects wording, spacing, or text placement.

World-aware prompts

Some prompts need more than style words. Nano Banana 3 can support product explainers, educational visuals, structured layouts, historical or geographic scenes, and diagrams when you give enough context.

For grounded images, mention the subject, era, location, real-world constraints, and the teaching or communication goal.

Multi-reference composition

You can combine references for people, products, poses, materials, environments, or style. This helps you build a coherent result from several inputs without rebuilding the whole concept from scratch.

Localized design ideas

Nano Banana 3 can adapt the same creative idea for different markets. Change the language, scene, cultural cues, product context, or visual format while keeping the core concept recognizable.

Prompt optimization

Prompt optimization helps turn rough notes into more useful image instructions. Use it when your idea is clear in your head but too short on the page.

A stronger prompt usually covers:

• subject
• setting
• composition
• camera and lens feel
• lighting
• color and material direction
• typography
• constraints and things to preserve

Who should use Nano Banana 3?

Nano Banana 3 is built for creators and teams who need practical speed with more control than a one-shot generator.

• Marketers can test ad directions, campaign variants, and social formats.
• Ecommerce teams can create product scenes, backgrounds, and listing visuals.
• Designers can explore layouts, typography, mockups, and early creative directions.
• Founders can make launch visuals, pitch graphics, thumbnails, and landing page assets.
• Education creators can build explainers, diagrams, and labeled visuals.
• Agencies can compare more creative options before client review.

Pricing and credits

Nano Banana 3 uses credits for paid generation and editing workflows. Credit cost can vary by selected model, tool, quality setting, and generation type. Higher-quality or more complex workflows may consume more credits than lightweight edits.

For current plan details, visit Pricing.

Privacy and safe use

Nano Banana 3 processes prompts, uploads, references, and outputs to provide the service. The product states that your content is not used to train models.

Use the workspace responsibly:

• upload only material you have the right to use
• avoid illegal, unsafe, infringing, or abusive content
• review images before commercial use
• disclose AI assistance when law, client policy, or platform rules require it

Read the full Privacy Policy and Terms of Service for details.

FAQ

What is Nano Banana 3?
Nano Banana 3 is an online AI image workspace for generating and editing visuals from prompts and reference images. It is designed for fast concepting, readable design text, localized visuals, product mockups, and controlled image-to-image edits.

Is Nano Banana 3 the same as Nano Banana Pro?
This site uses Nano Banana 3 for the homepage workflow. Official Google references describe Nano Banana Pro as Gemini 3 Pro Image, a model for advanced image generation and editing. Model availability can vary by provider, account, and selected generator option.

What can I create with Nano Banana 3?
You can create ads, posters, thumbnails, product scenes, ecommerce visuals, storyboards, diagrams, localized graphics, mockups, and social content.

How do I get better text inside generated images?
Keep text short, quote the exact words, describe placement and type style, and avoid too many separate labels in one image. If the image is close, make a focused edit for the wording.

Can I edit existing images with references?
Yes. Upload a source image or references, then explain what should stay consistent and what should change.

Are Nano Banana 3 images ready to publish?
Review every image before publishing. Check text, brand fit, rights, factual accuracy, and platform requirements.

Start creating

Open Nano Banana 3, choose Text to Image or Image to Image, add your prompt and references, then generate, refine, review, and export.