Up to sixteen references
GPT Image models blend up to sixteen images; Nano Banana models take eight to fourteen.
Nano Banana 2.1 is live — sharper edits, cleaner text, up to 4K
Image to Image AI
Upload a picture and prompt the change — restyle, re-light, extend, or composite into something new.
AI Model 5 credits
Prompt0/2000
Avoid copyrighted characters, logos, music and sensitive or explicit content — it may be blocked.
Resolution
Aspect Ratio
Batch generation1
Estimated cost: 5 credits
Your creation appears here
Pick a model, describe the shot, and hit Generate.
Why Image to Image AI
Why creators choose Image to Image AI
Image to image changes the picture you already have, instead of starting from a blank canvas.
Keeps your subject
References anchor identity while the rest changes.
Plain-language edits
Describe the change and run it.
Up to 4K
Export edits at print-ready size.
Image to image guide
How to edit and restyle images with image to image
Image to image starts from pictures you upload and changes them according to your prompt. It covers restyling, background swaps, product reshoots, extending a scene, and combining several images. This page explains how to prepare references and describe changes.
What image to image can do
Typical edits include style transfer, background and lighting changes, seasonal variants, outpainting, and multi-image compositing. Nano Banana 2 accepts up to fourteen reference images, Nano Banana 2.1 and Lite up to ten, Nano Banana Pro up to eight, and the GPT Image models up to sixteen. Results export at up to 4K.
Writing edit prompts
Describe only what should change, and say what must stay. "Keep the bottle and label exactly as they are, place it on a sunlit kitchen counter" tells the model what to preserve. When you upload several images, assign each a role: product, scene, palette, or style.
Change one variable at a time. If an edit is nearly right, repeat the request with a single correction rather than rewriting everything.
Preserve
State what must remain unchanged.
Change
Describe the new scene, style, or light.
Assign roles
Say what each reference contributes.
Preparing source images
Use sharp, well-lit images with the subject clearly visible. Avoid mixing references that disagree on perspective or lighting, since the model has to reconcile them. For the strongest subject fidelity, supply one or two clean views of the product or person.
How to use Image to Image AI
How to use Image to Image AI in three steps
Add one or more sharp references.
00:0000:1500:30
Who uses Image to Image AI
Who uses Image to Image AI for image work
Photo retouchers
Background and lighting changes without masking.
Campaign designers
Adapt one image for many placements.
E-commerce teams
Reshoot catalog photos in new scenes.
Creators
Restyle portraits, art, and screenshots.
More models
Image to Image AI and every other model in one studio
Switch engines without switching tools — Nano Banana 2.1 plus the strongest public image and video models, side by side in the Nano Banana AI studio.
Multi-reference editingCharacter consistencyText rendering1K–4K outputConversational editsStyle transferVideo generationBatch runs
Muse VideoNewFlagship multimodal video with native audio
Seedance 2.5HotMulti-reference Seedance with native audioSeedance 2.0Previous-gen Seedance up to 4KSeedance 2.0 FastFastFastest Seedance tier for drafts
MiniMax H3Fast text/image to video up to 2KMiniMax H3 MaxHotReference-driven video consistency
Veo 3.1Google’s cinematic drafting engine
Wan 3.0Flexible duration and adaptive framing
Kling 3.0Dramatic motion from a single prompt
GPT Image 2.5NewSharp text and precise edits to 4KGPT Image 2Versatile image generation to 4KNano Banana 2.1NewFast drafts and precise edits to 4KNano Banana ProHotHigh-fidelity image output to 4KNano Banana 2Fast drafts and conversational editsNano Banana 2 LiteFastLightweight tier for quick, low-cost draftsFAQ
Image to Image AI FAQ
What can image-to-image do?
Style transfer, background swaps, product reshoots, outpainting, and multi-image compositing.
How many images can I upload?
Up to eight references on the Nano Banana models and four on GPT Image models.
Will it keep my product accurate?
References anchor identity — describe only what should change.
What resolution do edits export at?
Up to 4K.