Prompt-generated images include the Wan 3.0 site mark when removal is not available on your account.
View paid options for watermark controlYour first prompt result will appear here. Start with one clear visual direction, then compare later versions in My Creations.
Fast image generation and editing
Nano Banana 2 is built for text-to-image and image editing workflows that need quick iteration without losing practical control. It fits campaign visuals, diagrams, product scenes, localized layouts, and reference-led edits.
Use this page to understand the model's documented strengths, prepare a tighter prompt, and move into the current Wan 3.0 image workspace with a clear review checklist.
External resource: nano banana 2
Core capabilities
Six production-focused strengths cover rapid iteration, flexible resolution, multilingual text, current-context grounding, multi-reference consistency, and precise edits.
01 / SPEEDBuilt for fast text-to-image generation and repeated edit passes, the model helps teams compare compositions, campaign directions, and product treatments without slowing early-stage review.
Start a new image
02 / OUTPUTChoose 0.5K, 1K, 2K, or 4K output and work across standard, portrait, landscape, and ultra-wide formats. This makes one workflow useful for quick drafts, social creative, product pages, and detailed final review.
Plan the output
03 / TEXTImproved multilingual text rendering supports posters, packaging concepts, diagrams, menus, and localized campaign layouts. Supply the final wording and language so typography becomes part of the composition from the first draft.
Open text to image
04 / CONTEXTWhen grounding is enabled, text and image search can inform scenes with current information and more specific real-world context. It is useful for location-led concepts, timely infographics, and visuals built around recognizable objects.
Read grounding details
05 / REFERENCESThe model can combine references for up to ten objects and preserve resemblance for up to four characters in one workflow. Assign each image a clear role to keep products, people, materials, and style cues distinct.
Open reference editing
06 / EDITSAdd, remove, restyle, relight, or reposition selected elements through follow-up instructions while preserving the parts that already work. Multi-turn editing is especially useful for product variations and localized design updates.
Start an image editPractical workflow
Start with one clear image objective, give every reference a defined role, and revise against a specific visual target.
Define the subject, scene, intended channel, aspect ratio, resolution, and exact text. For edits, state what must remain unchanged before describing the requested change.
Write the prompt in a clear order: subject, composition, environment, light, material, text, and output format. Label references by purpose, such as character, product, pose, style, or background.
Check composition, identity, product geometry, text, reflections, edges, and crop safety. Use a focused follow-up instruction for each revision instead of rewriting the entire prompt.
Workflow selection
Match the input and review criteria to the image task before writing a full production prompt.
| Workflow | Text to image | Reference-led editing | Text-rich visuals |
|---|---|---|---|
| Best input | A structured text brief with subject, scene, composition, lighting, and format. | One or more source images with explicit preserve and change instructions. | Final copy, target language, visual hierarchy, and intended layout. |
| Core task | Create campaign concepts, product scenes, illustrations, and editorial visuals. | Change backgrounds, styling, lighting, objects, or composition while retaining key details. | Build posters, packaging concepts, diagrams, menus, and localized creative. |
| Useful strength | Fast iteration, broad visual knowledge, flexible ratios, and high-resolution output. | Multi-reference consistency and conversational, instruction-led refinement. | Improved text rendering, multilingual layouts, and grounded information when enabled. |
| Review focus | Composition, realism, object count, materials, and final crop. | Identity, product shape, unchanged regions, shadows, and edge continuity. | Spelling, numerals, line breaks, hierarchy, and language accuracy. |
Production use cases
Use these capability paths to move from model research to a specific creative task.
Campaign development
Global creative
Controlled revision
Connected creation paths
Create a new still, revise an existing image, or carry an approved frame into motion.
Create from a brief
Revise a source
Carry the frame into motion
Frequently asked questions
Direct answers about image generation, editing, output sizes, reference handling, text rendering, and grounding.
Nano Banana 2 is an image generation and editing model designed for fast iteration and high-quality output. Its core strengths include 4K generation, multilingual text rendering, real-world context, multi-reference consistency, and conversational editing.
Yes. It supports text-to-image generation, image-guided creation, and multi-turn editing. You can continue refining an existing result to change language, layout, lighting, styling, or selected objects without rebuilding the image from scratch.
The model supports 0.5K, 1K, 2K, and 4K output. Alongside common square, portrait, and landscape formats, it supports ultra-wide ratios such as 4:1 and 8:1 for banners, headers, and panoramic layouts.
It is well suited to text-rich visuals such as posters, packaging concepts, diagrams, menus, and localized marketing assets. Provide the exact wording and target language, then verify spelling, numbers, line breaks, and trademarks before publishing.
A single workflow can use references for up to ten objects and maintain resemblance for up to four characters. Clear role labels and compatible source images make it easier to preserve identity, product shape, materials, and style.
When enabled, search grounding gives the model current text and image context for real places, objects, events, weather, and data-led visuals. It improves specificity, but factual details and usage rights should still be reviewed before publication.
The model can work with text, images, video, and PDF context, then return image and text output. This supports workflows such as turning a video into a poster concept or using a document as context for a visual summary.
It is a strong fit for campaign concepting, product scenes, editorial graphics, localized marketing assets, diagrams, social creative, and reference-led image edits that benefit from fast iteration and flexible output sizes.
Review the full-size image for text accuracy, identity, product shape, object count, hands, material behavior, lighting consistency, crop safety, and rights for every input. Rebuild critical copy and compliance text in editable design layers when the image is destined for a brand or commercial page.
From concept to production
Use Wan 3.0 image tools to create a new composition, revise source material, and prepare an approved still for the next stage of production.
Start from text for a new composition or upload a source image for a controlled revision.
Define the aspect ratio, resolution, language, and final placement before generating.
Keep approved details stable and use focused instructions for each visual change.
Use the finished still in a campaign, product page, localized layout, or image-to-video workflow.
Prompt-generated images include the Wan 3.0 site mark when removal is not available on your account.
View paid options for watermark controlYour first prompt result will appear here. Start with one clear visual direction, then compare later versions in My Creations.
Fast image generation and editing
Nano Banana 2 is built for text-to-image and image editing workflows that need quick iteration without losing practical control. It fits campaign visuals, diagrams, product scenes, localized layouts, and reference-led edits.
Use this page to understand the model's documented strengths, prepare a tighter prompt, and move into the current Wan 3.0 image workspace with a clear review checklist.
External resource: nano banana 2
Core capabilities
Six production-focused strengths cover rapid iteration, flexible resolution, multilingual text, current-context grounding, multi-reference consistency, and precise edits.
01 / SPEEDBuilt for fast text-to-image generation and repeated edit passes, the model helps teams compare compositions, campaign directions, and product treatments without slowing early-stage review.
Start a new image02 / OUTPUTChoose 0.5K, 1K, 2K, or 4K output and work across standard, portrait, landscape, and ultra-wide formats. This makes one workflow useful for quick drafts, social creative, product pages, and detailed final review.
Plan the output03 / TEXTImproved multilingual text rendering supports posters, packaging concepts, diagrams, menus, and localized campaign layouts. Supply the final wording and language so typography becomes part of the composition from the first draft.
Open text to image04 / CONTEXTWhen grounding is enabled, text and image search can inform scenes with current information and more specific real-world context. It is useful for location-led concepts, timely infographics, and visuals built around recognizable objects.
Read grounding details05 / REFERENCESThe model can combine references for up to ten objects and preserve resemblance for up to four characters in one workflow. Assign each image a clear role to keep products, people, materials, and style cues distinct.
Open reference editing06 / EDITSAdd, remove, restyle, relight, or reposition selected elements through follow-up instructions while preserving the parts that already work. Multi-turn editing is especially useful for product variations and localized design updates.
Start an image editPractical workflow
Start with one clear image objective, give every reference a defined role, and revise against a specific visual target.
Define the subject, scene, intended channel, aspect ratio, resolution, and exact text. For edits, state what must remain unchanged before describing the requested change.
Write the prompt in a clear order: subject, composition, environment, light, material, text, and output format. Label references by purpose, such as character, product, pose, style, or background.
Check composition, identity, product geometry, text, reflections, edges, and crop safety. Use a focused follow-up instruction for each revision instead of rewriting the entire prompt.
Workflow selection
Match the input and review criteria to the image task before writing a full production prompt.
| Workflow | Text to image | Reference-led editing | Text-rich visuals |
|---|---|---|---|
| Best input | A structured text brief with subject, scene, composition, lighting, and format. | One or more source images with explicit preserve and change instructions. | Final copy, target language, visual hierarchy, and intended layout. |
| Core task | Create campaign concepts, product scenes, illustrations, and editorial visuals. | Change backgrounds, styling, lighting, objects, or composition while retaining key details. | Build posters, packaging concepts, diagrams, menus, and localized creative. |
| Useful strength | Fast iteration, broad visual knowledge, flexible ratios, and high-resolution output. | Multi-reference consistency and conversational, instruction-led refinement. | Improved text rendering, multilingual layouts, and grounded information when enabled. |
| Review focus | Composition, realism, object count, materials, and final crop. | Identity, product shape, unchanged regions, shadows, and edge continuity. | Spelling, numerals, line breaks, hierarchy, and language accuracy. |
Production use cases
Use these capability paths to move from model research to a specific creative task.
Campaign development
Global creative
Controlled revision
Connected creation paths
Create a new still, revise an existing image, or carry an approved frame into motion.
Create from a brief
Revise a source
Carry the frame into motion
Frequently asked questions
Direct answers about image generation, editing, output sizes, reference handling, text rendering, and grounding.
Nano Banana 2 is an image generation and editing model designed for fast iteration and high-quality output. Its core strengths include 4K generation, multilingual text rendering, real-world context, multi-reference consistency, and conversational editing.
Yes. It supports text-to-image generation, image-guided creation, and multi-turn editing. You can continue refining an existing result to change language, layout, lighting, styling, or selected objects without rebuilding the image from scratch.
The model supports 0.5K, 1K, 2K, and 4K output. Alongside common square, portrait, and landscape formats, it supports ultra-wide ratios such as 4:1 and 8:1 for banners, headers, and panoramic layouts.
It is well suited to text-rich visuals such as posters, packaging concepts, diagrams, menus, and localized marketing assets. Provide the exact wording and target language, then verify spelling, numbers, line breaks, and trademarks before publishing.
A single workflow can use references for up to ten objects and maintain resemblance for up to four characters. Clear role labels and compatible source images make it easier to preserve identity, product shape, materials, and style.
When enabled, search grounding gives the model current text and image context for real places, objects, events, weather, and data-led visuals. It improves specificity, but factual details and usage rights should still be reviewed before publication.
The model can work with text, images, video, and PDF context, then return image and text output. This supports workflows such as turning a video into a poster concept or using a document as context for a visual summary.
It is a strong fit for campaign concepting, product scenes, editorial graphics, localized marketing assets, diagrams, social creative, and reference-led image edits that benefit from fast iteration and flexible output sizes.
Review the full-size image for text accuracy, identity, product shape, object count, hands, material behavior, lighting consistency, crop safety, and rights for every input. Rebuild critical copy and compliance text in editable design layers when the image is destined for a brand or commercial page.
From concept to production
Use Wan 3.0 image tools to create a new composition, revise source material, and prepare an approved still for the next stage of production.
Start from text for a new composition or upload a source image for a controlled revision.
Define the aspect ratio, resolution, language, and final placement before generating.
Keep approved details stable and use focused instructions for each visual change.
Use the finished still in a campaign, product page, localized layout, or image-to-video workflow.
