gift

Black Friday: limited-time 50% off!

Claim

Your image. A new direction.

Image to Image AI Generator

Start with a photo, sketch or design and tell Nano Banana what to change. Restyle an image, replace its background or explore a new variation using your original as a reference. Upload your image in the workspace below.

Nano Banana AI Creation Workbench

Get 20 Free Credits & Try Pro

Compare →

Reference Images

Describe your idea

0/20000
Detailed descriptions lead to better results
Credit Cost
8credits
New Release! Nano Banana Pro
Real-world Knowledge

Real-world Knowledge

Use the Gemini model’s real-world knowledge and deep reasoning capabilities to deliver precise, detailed, rich image results. Annotate pictures, represent data as infographics, or turn handwritten notes into diagrams.

Clear Text Generation

Clear Text Generation

Sharp, legible text helps you create impactful posters, intricate diagrams, and detailed product mockups. Describe what type of font you want, or simulate different types of handwriting.

Perfect Style Transfer

Perfect Style Transfer

Precisely capture and transfer artistic styles—whether oil painting, watercolor, illustration, or photography—flawlessly replicate and apply to your creations.

4K Ultra Resolution

4K Ultra Resolution

Support up to 4K ultra-high resolution output with every detail crystal clear, meeting professional printing and commercial use requirements.

Create from a reference

What is image to image AI?

Image to image, also called img2img, creates a new image from an existing picture and a text prompt. Your reference supplies visual context, such as the subject, layout or colors; your prompt describes the changes. It is useful when you already have a starting image and want to take it further.

Use the Nano Banana workspace for both text to image and image editing. On this page, image to image is selected so you can begin with a reference.

One image, more creative possibilities

Use an AI image generator from an image to explore a specific change while giving the model a visual starting point.

01

Turn a photo into a new art style

Try watercolor, editorial illustration, anime or a cinematic treatment. Name the style and the details you want to keep, such as the subject's pose, clothing or the original composition.

02

Reimagine product backgrounds

Place a product in a studio scene, on a tabletop or in a lifestyle setting. Ask for matching light and shadows, and specify that the product's shape, color and label should stay consistent.

03

Develop sketches and design ideas

Use a sketch or rough concept as a reference for a more finished visual. Describe the materials, lighting and level of realism you need, then compare the result with your original layout.

04

Create variations and keep refining

Explore a new palette, setting or mood, then use Continue Edit on a result to make another change. Working through one adjustment at a time makes it easier to see what each prompt does.

How to use image to image

Upload a reference, explain the edit and review your result in the Nano Banana workspace.

1

Upload your reference image

Keep Image to Image selected and upload a JPG, PNG or WebP file. Start with a clear image where the subject is easy to see. The selected model determines the available reference-image and output options.

2

Describe what to change and keep

Write a specific instruction: change the background, restyle the scene or adjust an object. Also name the details to preserve. Choose your model and any available size or resolution settings.

3

Generate, review and download

Sign in, check the displayed credit cost and generate. Review the subject, text and edges before downloading. Use Continue Edit if you want to refine the result with another prompt.

Image to image prompt examples

Use this structure: keep the important details, change one thing, then describe the finish. Adapt these prompts to your uploaded image, or browse more Nano Banana prompts.

Product photography

Give a product a new setting

Keep the bottle's shape, cap, color and label from the reference. Replace the background with a pale stone tabletop beside a window. Add soft daylight from the left and a realistic contact shadow. Keep the camera angle unchanged.

Photo to art

Restyle a portrait

Turn this portrait into a watercolor illustration on textured paper. Keep the person's hairstyle, clothing, expression and pose recognizable. Use a muted teal and warm neutral palette, with a simple background and soft brushwork.

Sketch to image

Develop a room concept

Use this room sketch as the composition reference. Keep the window, sofa and table in their original positions. Create a realistic interior with oak flooring, linen upholstery and soft morning light. Do not add extra furniture.

Get closer to the edit you have in mind

A useful reference and a focused prompt make it easier to judge and improve each result.

Start with the clearest source

Choose a sharp, well-lit image and avoid heavy crops that hide the subject. If the source is blurry, try unblurring the image before making creative edits.

Be explicit about what stays

Instead of only saying 'make it cinematic,' specify the lighting change and ask to preserve the camera angle, expression and clothing. Review faces, logos and small text carefully; a generated edit can alter them.

Separate editing from finishing

Use image to image for visual changes. For a transparent cutout, use the background remover. To increase the size of a finished image, use the image upscaler.

Simple and transparent Nano Banana pricing

Choose the perfect plan for your AI image creation needs.

Cancel anytime

Starter

Save 30%
$10.49
$14.99USD / mo
Billed annually · $126.00 / Year
100 credits ≈ $3.00

Perfect for trying out AI image generation and small projects.

What's included

  • 4200 AI generation credits per year
  • Nano Banana models
  • Nano Banana Pro models
  • Seedream 4.5 models
  • Veo 3.1 & Sora 2 & Wan 2.5/2.6 video generation models
  • Flux Kontext models
  • Up to 5 reference images upload
  • Up to 2 batch generation tasks
  • Nano Banana Pro up to 2K resolution
  • 3x,4x image enhancement service
  • Standard generation speed
  • Standard success rate
  • Standard customer support
  • Commercial license

Perfect for experiencing AI image generation and editing effects.

Most Popular

Pro

Save 50%
$19.99
$39.99USD / mo
Billed annually · $239.94 / Year
100 credits ≈ $1.25

Ideal for creators, designers, and content professionals.

Everything in Starter

  • 19200 AI generation credits per year
  • Nano Banana models
  • Nano Banana Pro models
  • Seedream 4.5 models
  • Veo 3.1 & Sora 2 & Wan 2.5/2.6 video generation models
  • Unlimited Flux Kontext models
  • Up to 10 reference images upload
  • Up to 10 batch generation tasks
  • Nano Banana Pro up to 4K resolution
  • 3x,4x image enhancement service
  • Priority processing speed
  • High success rate
  • High-speed download channel
  • Priority customer support
  • Unlimited commercial license

🔥 Limited Time Offer! Get Pro Now

Premium

Save 50%
$59.5
$119USD / mo
Billed annually · $714.00 / Year
66,000 credits
Billed annually
100 credits ≈ $1.08

For teams, agencies, and high-volume commercial use.

Everything in Pro

  • 66,000 credits / Year
  • Nano Banana models
  • Nano Banana Pro models
  • Seedream 4.5 models
  • Veo 3.1 & Sora 2 & Wan 2.5/2.6 video generation models
  • Unlimited Flux Kontext models
  • Up to 10 batch generation tasks
  • Up to 10 reference images upload
  • Nano Banana Pro up to 4K resolution
  • 3x,4x image enhancement service
  • Fastest generation speed
  • Most advanced AI editing capabilities
  • Permanent image storage
  • Dedicated account manager
  • Unlimited commercial license

Perfect for teams and businesses

For any questions about payment, contact our support team:Contact support

Pay safely and securely with

MastercardVisaAmericanExpressApplePayUnionPayGooglePayJCBDiscoverClickToPay

Image to image FAQ

Practical answers about references, prompts, credits and editing results.

What is the difference between image to image and text to image?

Text to image starts with a written description. Image to image adds an existing picture as a visual reference. Choose text to image for a new concept from scratch, and image to image when you want to change or develop a photo, sketch or design you already have.

How do I generate an AI image from an image?

Upload a reference in the workspace, write what you want to change and choose a model. Sign in and review the credit cost before generating. When the result is ready, download it or select Continue Edit to use it as the next reference.

Can I use image to image for free?

Generation uses credits. If your account has free credits available, you can use them for supported image tasks. The cost depends on the selected model, resolution and output count and is shown before generation. See pricing for current plans and credits.

Can I keep the same face, product or composition?

A clear reference and explicit instructions can help preserve recognizable details. State what must stay the same and keep each edit focused. Exact identity, lettering and geometry are not guaranteed, so compare the output with the original before using it.

Can I turn a photo into anime, a painting or another style?

Yes. Upload the photo and describe the visual treatment you want, such as anime, watercolor or an editorial illustration. Include the pose, objects and composition you want to keep so the prompt covers both the new style and the original content.

Which image formats and reference counts are supported?

The workspace accepts JPG, PNG and WebP images. Reference-image counts and file-size limits depend on the selected model; check the upload area for its current limits. If using several references, explain each image's role in the prompt.

How do I control how much the original image changes?

Use the prompt to set the scope: name one change and list the details to preserve. For a larger transformation, describe the new style or scene more fully. This workspace uses natural-language editing; it does not provide a Stable Diffusion denoising-strength slider.

Why did the AI change a detail I wanted to keep?

Image to image generates a new result from your reference rather than locking every original pixel. Simplify the prompt, remove conflicting instructions and explicitly name the detail to preserve. Try one edit at a time, using the original again if successive edits have drifted too far.

Start with what you have

Take your image in a new direction

Bring a photo, a sketch or a product shot. Describe one change and explore what it can become with Nano Banana.

Try image to image
Image to Image AI Generator | Nano Banana