Nano Banana Image Generator
Create from a prompt or refine a reference image in GenImager’s image workspace. Available models can change with your account and plan.
- Text-to-image workflow
- Reference image editing
- Plan limits apply
Nano Banana Prompt
Write a prompt or describe a reference edit
Image Preview
Review the generated image before downloading
Add a prompt or reference image to begin
Create or edit images with Nano Banana on GenImager
Nano Banana is Google DeepMind’s name for Gemini image generation and editing models. This independent GenImager page provides prompt guidance and the same text and reference image workflows available on GenImager. Generated images use GenImager’s active model, which may change; this is not Google or the Gemini app.
Start with a text prompt
- Choose Text to Image and describe the scene in 10 to 2,000 characters.
- Name the subject, composition, background, and lighting; include text only when it is important to the image.
- Generate and review the square output in the gallery before downloading.
Make a focused edit with a reference
- Choose Image to Image and upload up to two JPG, PNG, or WebP references, no larger than 10 MB each.
- State what to change, then list the subject details, colors, or framing to preserve.
- Review the result closely. If a detail changes unexpectedly, make one specific prompt revision and try again.
Prompt example: preserve the subject while changing the scene
Use this as a starting point. Adjust the subject and instructions for your own image.
Use the uploaded backpack as the reference. Keep its shape, pocket layout, zipper positions, and dark green fabric. Place it on a wooden bench beside a trail at soft morning light. Show the full backpack in a three-quarter view. Do not add a logo or change the product design.
Review the result and refine one detail at a time
- Use a clear, original reference image; avoid screenshots with borders or compression artifacts.
- Describe the role of each image when you upload two references.
- Inspect product shape, logos, text, and other details before using an edit in a listing or advertisement.
- The active model and plan allowance can change; check the current pricing and workflow before generating.
Questions about this image generator
Is GenImager an official Google or Gemini product?
No. GenImager is an independent service. Nano Banana is Google DeepMind’s model name; this page is not the Gemini app or a Google service.
Will this page always use a Nano Banana model?
No. This page uses GenImager’s active text-to-image and image-to-image workflows. The active model depends on GenImager’s configuration and can change.
Can I choose Nano Banana Pro or another version?
No model version selector is available here. Generation uses the active model configured for GenImager’s image workflows.
What files can I use for image editing?
Image-to-image accepts one or two JPG, PNG, or WebP files up to 10 MB each. The uploaded image is sent to the configured generation service for processing.