Character portraits
Describe an adult fictional character and a clear visual style.
Open the creation workspaceActual creation examples
Nano Banana 2: character scenes from a reference
Nano Banana 2 · Upload a product photo or portrait and explore new backgrounds, outfits and lighting. Compare the reference with actual edits, then apply the same approach to your own image.
Choose a model and settings before submitting; examples never auto-generate.
What would you like to create?
Find your starting point
Use cases
Choose a useful direction, then adapt the example to your own brief.
Everyday scenes
Use one portrait reference to explore a new location or activity.
Open the creation workspaceStory concepts
Develop character atmosphere and composition before final illustration.
Open the creation workspaceInput advice and practical limits
Start with a clear portrait, describe the new outfit, setting and props, and specify the facial details to retain. The pottery-workshop example shows one reference-led character scene. Check the face and hands before developing a series; each new generation can vary even when the prompt and reference stay the same.
Real outputs and practical use cases
These examples were generated through this site’s models or effects. Reference people and products are AI-created. Each is one result, not a fixed-output promise; prompts show the exact original wording. A button restores settings for review and never submits or spends credits automatically.
Actual creation examples
A portrait becomes a workshop scene
Use the same fictional portrait while changing clothing, setting and props. Useful for character lifestyle scenes and editorial concepts; inspect face and hand details.
896 × 1200 px
View the exact prompt
Keep the same fictional adult woman in the reference: short black bob, subtle freckles, warm smile. Change the scene to a small ceramic workshop. She wears a simple dark teal apron over her cream shirt, holding a small clay bowl, soft window light, blurred shelves of pottery in the background. Chest-up natural editorial portrait, realistic skin and fabric, no text or logos.
How to use this service
Step 01
Bring a clearer brief
Write the character’s distinctive appearance, clothing and setting.
Step 02
Select the input and settings
Use one portrait as a reference if you want a starting point for continuity.
Step 03
Review, save and reuse
Compare faces, hands and clothing across versions before choosing one.
Write a more useful creative brief
Keep a clear subject and a single goal. The example prompt is a starting point you can edit.
Subject + purpose + composition + light or motion + what to preserve
Open the creation workspaceSubject and input
Repeat a few distinctive features rather than a long list of competing instructions.
Composition and control
Keep outfit or background changes small when checking face continuity.
Review and refine
Treat each scene as a new generated image and inspect identity rather than assuming it is fixed.
Available creation controls
Review current options and credits in the workspace before submitting.
| Setting | Available options | How to choose |
|---|---|---|
| Input modes | Text to image / One image reference | Review current options and credits in the workspace before submitting. |
| Output size | 1K / 2K / 4K | Review current options and credits in the workspace before submitting. |
| Aspect ratios | auto / 1:1 / 16:9 / 9:16 / 4:3 / 3:4 / 3:2 / 2:3 | Review current options and credits in the workspace before submitting. |
| Prompt limit | 4000 characters | Review current options and credits in the workspace before submitting. |
Choose the output for its destination
Set the ratio and available output options in the workspace. Inspect the actual result before export. Choose only ratios supported by the selected model. A new ratio can reframe or regenerate the image.
| Format | Suitable use | Practical advice |
|---|---|---|
| 1:1 | Product posts and compact social layouts | Keep the main subject central and leave a comfortable margin. |
| 2:3 / 3:4 / 9:16 | Portraits, vertical posters and mobile feeds | Plan headroom and text placement; avoid cutting important details. |
| 3:2 / 16:9 | Website covers, concept art and widescreen video | Reserve side space and keep the focal subject readable at small sizes. |
Continue with a related task
Build the next part of your project using the matching workflow.
AI Image Editor
Upload a product photo or portrait and explore new backgrounds, outfits and lighting. Compare the reference with actual edits, then apply the same approach to your own image.
AI Image Enhancer
Explore detail, noise and exposure adjustments for portraits and product visuals. Compare the source and result, keep natural texture and decide whether the edit is useful.
Image to Video AI
Turn a product photo, portrait or concept image into a short clip. Direct restrained subject motion and camera movement for product showcases and social drafts.
Text to Image AI
Make event posters, social covers, editorial illustrations and cinematic concepts from a written brief. Explore actual outputs and reuse a prompt to start.
Questions before creating
What edits is this workflow suited to?
Product scene changes, portrait styling and campaign drafts. Generative editing redraws the image and cannot promise pixel-exact preservation. Review logos, measurements and important details in an editor; there is no brush-mask workflow in this workspace.
How many credits does generation cost?
Cost depends on the model and output settings and is shown before submission. Check a direction at a smaller setting first; higher resolution is not a quality guarantee.
Can I reuse these examples?
Yes. The button restores the actual model, prompt and reusable reference. Replace it with your own material if you prefer; your generated result can differ from the example.
Can I use my own photo instead of the example?
Yes. Replace the reference in the workspace before submitting. Use an image you can use for the project; one image is supported by the current image-editing workflow.
Can I create from a phone?
You can open the workspace in a mobile browser. Prepare the original image first and review the result at its full size before publishing.
Where can I save the finished result?
After a successful generation, open the result or creation history to save it. Inspect the original file and add exact text or finishing edits where needed.
Create your own version
Choose a model and settings before submitting; examples never auto-generate.
Start creating