Available on this page
Text-to-image
Describe the subject, setting, composition, lighting, and constraints in a prompt.
AI Image Generator
Nano Banana 2 AI Image Generator
Nano Banana 2 is a common name for Google Gemini 3.1 Flash Image. It is a Gemini image generation and editing model that can create images from text and use reference images for editing, recomposition, and visual extension.
Direct answer
Nano Banana 2 is a image model from Google available through Videmos. This page shows the configured inputs, output controls, credit cost of 2 per use, examples or workflow notes when available, and links to the provider information used to describe the model. Check the generator before submitting because availability follows the current Videmos configuration.
Last updated: July 12, 2026
text prompt / up to 4 reference images2 creditsGoogle
Create with Nano Banana 2 using the inputs and output settings shown above.
Model overview
What is Nano Banana 2?
Nano Banana 2 is a common name for Google Gemini 3.1 Flash Image. It is a Gemini image generation and editing model that can create images from text and use reference images for editing, recomposition, and visual extension.
A workflow can begin with a prompt or with existing product, character, layout, or scene images. Multiple inputs can help define subjects and environments, while this page keeps the current reference-image limit visible.
Credits, reference limits, aspect ratios, and resolutions shown here follow the current Videmos generator controls.
Creation options
Ways to create with Nano Banana 2
Review the available input methods and output choices before starting an image.
Model identity
- Official model identity: Gemini 3.1 Flash Image.
- This page is fixed to Nano Banana 2.
- The generator supports text-to-image and image-to-image entry points.
- Input, aspect ratio, and output choices are available directly in the generator.
Generator options
Use a text prompt, up to 4 reference images, 6 aspect ratios, and 1K / 2K / 4K output tiers for 2 credits per generation.
- Videmos input
- Text prompt + Up to 4 reference images
- Maximum references
- Up to 4 reference images
- Aspect ratio
- auto / 1:1 / 4:3 / 3:4 / 16:9 / 9:16
- Resolution / quality
- 1K / 2K / 4K
- Images per submission
- 1
- Credits
- 2 credits/use
Choose the input method, aspect ratio, and output tier, then start generating on this page.
Core capabilities
Six ways to generate and edit
These workflows combine prompts, reference images, aspect ratios, and output tiers for common image tasks.
Available on this page
Reference-image editing
Upload an image and describe changes to the setting, composition, material, or visual details.
Up to 4 images on this page
Multi-image composition
Assign a role to each visual reference and describe how the subjects should relate.
Can be attempted here
Text layout
Specify exact copy, language, typographic character, hierarchy, and placement for covers or infographics.
6 ratios on this page
Aspect-ratio control
Choose a ratio for social posts, covers, posters, or banners.
1K / 2K / 4K on this page
Resolution output
Choose an exposed resolution for composition review or delivery needs.
适合谁使用
Nano Banana 2 更适合这些图片工作流
这些使用建议围绕提示词、参考图、画幅和输出档位展开，帮助不同团队判断适合自己的创作方式。
01灵感到草图
个人创作者
快速把灵感、角色、封面或社媒视觉从文字变成可评估的图片草案。
02商品视觉
电商团队
用商品参考图测试背景、布光、道具和横竖版主图，不必每次重新搭建拍摄场景。
03素材迭代
广告素材团队
围绕同一卖点生成不同构图和视觉方向，再筛选适合投放测试的版本。
04多画幅
社媒运营
按平台比例准备方图、横图和竖图，用统一提示词维持账号视觉调性。
05概念验证
设计与概念团队
把草图、情绪板或参考图转成更完整的视觉方向，用于内部评审和概念预演。
06内容配图
内容团队
为文章、专题、课程或活动页准备封面、插图和信息图方向，并保留来源可核验案例。
Sourced examples
Traceable Nano Banana 2 image examples
Images come from Google documentation. Videmos adapts each prompt as a concise reference for trying a similar composition; use the source links to inspect the originals.
Google image generation documentation
01
Minimal magazine cover
Tests specified title copy, date, barcode, and foreground-background hierarchy.
- Model identity
- Gemini 3.1 Flash Image
- Mode
- Text-to-image
- Suggested settings
- 3:4 · 2K
Reference prompt
A minimal blue magazine displayed on a designer-store shelf. Large serif title Nano Banana, a person in a sleek dress holding the number 2, issue number, Feb 2026, and a barcode in the lower corner. Orange plaster wall, no text beyond the requested copy.
Google image generation documentation
02
Isometric garden and number-shaped pool
Tests strict viewpoint, architecture, materials, and integrated lettering.
- Model identity
- Gemini 3.1 Flash Image
- Mode
- Text-to-image
- Suggested settings
- 16:9 · 2K
Reference prompt
A perfectly isometric photograph of a modern garden with a number 2-shaped pool, pale stone paths, trimmed plants, and a minimal residence. Clear white dimensional lettering reads Nano Banana 2 beside the water, natural daylight, clean composition.
Google image generation documentation
03
Circular coffee-shop logo
Tests specified brand copy, typography, palette, and logo composition.
- Model identity
- Gemini 3.1 Flash Image
- Mode
- Text-to-image
- Suggested settings
- 1:1 · 1K
Reference prompt
Create a modern, minimalist logo for a coffee shop called The Daily Grind. Use a clean, bold sans-serif font, a black-and-white palette, a circular composition, and integrate a coffee bean in a clever way.
Google image generation documentation
04
Black ceramic mug product photo
Tests material, studio lighting, camera angle, and a square product composition.
- Model identity
- Gemini 3.1 Flash Image
- Mode
- Text-to-image
- Suggested settings
- 1:1 · 1K
Reference prompt
A high-resolution studio product photograph of a minimalist matte-black ceramic coffee mug on polished concrete. Use three-point softbox lighting, soft highlights without harsh shadows, a slightly elevated 45-degree camera angle, sharp steam detail, and a square composition.
Google image generation documentation
05
Minimal red maple composition
Tests subject placement, negative space, and soft lighting for a text-ready background.
- Model identity
- Gemini 3.1 Flash Image
- Mode
- Text-to-image
- Suggested settings
- 1:1 · 1K
Reference prompt
A minimalist composition with one delicate red maple leaf in the bottom-right. Use a vast clean off-white background with generous negative space for text, soft diffused light from the upper-left, and a square composition.
Prompting guide
Turn visual requirements into an executable brief
State required content first, then add photography, design, and exclusion constraints.
Subject + setting + composition + camera + lighting + exact text + constraints
Subject
Define people, products, animals, or objects, including count and identifying traits.
Setting
Name the place, time, weather, materials, and background elements.
Composition
Specify framing, placement, negative space, and foreground-to-background relationships.
Camera
Describe viewpoint, focal character, depth of field, motion, or isometric projection.
Lighting
Set light direction, softness, color temperature, and reflections.
Exact text
Write the precise copy, language, placement, and hierarchy that should appear.
Constraints
List unwanted copy, marks, people, elements, or composition errors.
Technical controls and limits
Choose the right creation settings
Use these specifications to confirm the input, aspect ratio, output tier, and credit cost for an image task.
Input and output
- Videmos input
- Text prompt + Up to 4 reference images
- Maximum references
- Up to 4 reference images
- Aspect ratio
- auto / 1:1 / 4:3 / 3:4 / 16:9 / 9:16
- Resolution / quality
- 1K / 2K / 4K
- Images per submission
- 1
- Credits
- 2 credits/use
Available workflows
- Text-to-image
- Image-to-image
- Up to 4 reference images
- auto / 1:1 / 4:3 / 3:4 / 16:9 / 9:16
- 1K / 2K / 4K
- 1 image per submission
Review small lettering, multi-person relationships, and fine edges after generation before continuing an edit.
Model comparison
Nano Banana 2 compared with three image models
Compare input methods, reference limits, aspect ratios, output controls, image count, and credits. The workflow suggestions are not an image-quality ranking.
- Official identity
- Google Gemini 3.1 Flash Image
- Videmos input
- Text prompt + 4 reference images
- Maximum references
- 4
- Aspect ratio
- auto / 1:1 / 4:3 / 3:4 / 16:9 / 9:16
- Resolution / quality
- 1K / 2K / 4K
- Images per submission
- 1
- Credits
- 2 credits/use
- Videmos workflow suggestion
- For one-shot image tasks using text or up to four references with 1K to 4K output controls.
- Official identity
- Google Gemini 3 Pro Image
- Videmos input
- Text prompt + 4 reference images
- Maximum references
- 4
- Aspect ratio
- auto / 1:1 / 4:3 / 3:4 / 16:9 / 9:16
- Resolution / quality
- 1K / 2K / 4K
- Images per submission
- 1
- Credits
- 3 credits/use
- Videmos workflow suggestion
- For a similar reference and ratio workflow using Nano Banana Pro.
- Official identity
- OpenAI GPT Image 1.5
- Videmos input
- Text prompt + 4 reference images
- Maximum references
- 4
- Aspect ratio
- 1:1 / 3:2 / 2:3
- Resolution / quality
- auto / high / medium / low
- Images per submission
- 1 / 2 / 4
- Credits
- 6 credits/use
- Videmos workflow suggestion
- For tasks that need 1, 2, or 4 outputs and GPT Image size and quality controls.
- Official identity
- ByteDance Seedream 5.0
- Videmos input
- Text prompt + 14 reference images
- Maximum references
- 14
- Aspect ratio
- auto / 1:1 / 2:3 / 3:2 / 3:4 / 4:3 / 4:5 / 5:4 / 9:16 / 16:9 / 21:9
- Resolution / quality
- 2K / 3K
- Images per submission
- 1 / 2 / 4
- Credits
- 5 credits/use
- Videmos workflow suggestion
- For tasks that need more references, more ratio choices, or 1, 2, or 4 outputs.
Start with reference limits and output controls, then consider the credits used by each generation.
How to use
How to use Nano Banana 2
Follow four steps from choosing an input method to reviewing the generated image.
1. Choose an input mode
Start with text-to-image or image-to-image; upload up to four references when needed.
2. Write the prompt
Define the subject, setting, composition, lighting, exact text, and constraints.
3. Set the output
Choose from the aspect ratios and 1K / 2K / 4K settings currently exposed.
4. Submit and review
Each submission uses the displayed 2 credits; inspect text, subject relationships, and edges after generation.
Frequently asked questions
Nano Banana 2 FAQ
What is the official model name for Nano Banana 2?+
Nano Banana 2 is a common name for Google Gemini 3.1 Flash Image, an image generation and editing model.
How many reference images can I use here?+
You can add up to four reference images and review or remove each one before submission.
How should I use multiple reference images?+
Assign each image a clear role such as subject, product, clothing, background, or layout, then explain those relationships in the prompt.
Which aspect ratios are available?+
Choose auto, 1:1, 4:3, 3:4, 16:9, or 9:16.
Which resolutions are available?+
Choose 1K, 2K, or 4K. The final result also depends on the input images, prompt, and composition complexity.
How many images are generated per submission?+
Each generation returns one image so you can refine the same prompt and references step by step.
Can it place text inside an image?+
You can specify exact copy, language, placement, and hierarchy, but spelling, characters, and layout should still be reviewed after generation.
How should I choose between Nano Banana 2 and Nano Banana Pro?+
Use the comparison table to review reference-image count, output choices, and credits for your image task.
How can I continue editing a result?+
Upload a previous result as a new reference image and submit another image task.
How many credits does one generation use?+
Each Nano Banana 2 generation uses 2 credits, shown beside the generation action.
Sources
Reference and example sources
These sources help verify the model identity, image examples, and prompting structure shown on the public page.