1 available
TextImage

Add images

Supports multiple images (up to 10)

Max 10MB per image

0 characters
Preview

Nano Banana by Google DeepMind

Nano Banana by Google DeepMind

What is Nano Banana?

Nano Banana is Google DeepMind's original Gemini image generation and editing model. It established the Nano Banana family with native multimodal editing, conversational prompting, and stronger likeness preservation in image revisions.

Google DeepMind builds the model, and Nano Banana is the original family base

Google DeepMind builds the model, and Nano Banana is the original family base

Nano Banana belongs to Google's Gemini image model family. Google DeepMind develops the model capabilities, while Google uses the family across its own products and creative experiences.

It introduced native multimodal image editing inside Gemini

It introduced native multimodal image editing inside Gemini

Nano Banana is designed to work from both text and image input. You can upload an image, describe the change in everyday language, and keep refining the same visual through conversational edits instead of restarting each time.

Its practical strengths are likeness preservation, photo blending, and multi-turn editing

Its practical strengths are likeness preservation, photo blending, and multi-turn editing

Google presented Nano Banana around realistic editing of people, pets, and existing photos. The model is strong when you need to preserve the subject, blend multiple photos into one scene, or keep editing the same image across multiple rounds.

Nano Banana is the foundation that Pro and 2 build on

Nano Banana is the foundation that Pro and 2 build on

Nano Banana Pro pushes the family toward studio-quality precision and control. Nano Banana 2 brings many of those stronger capabilities into a faster Flash model. Nano Banana remains the original Gemini-native generation and editing base inside the series.

How to use Nano Banana

Use Nano Banana when you want the original Gemini image workflow for generation, editing, and conversational refinement.

Start with text or upload an image

Start from a prompt when you need a new concept, or upload an existing photo when you want to change the background, outfit, scene, styling, or layout while keeping the strongest parts of the source image.

Describe the change in direct language

Nano Banana is built for conversational image editing. Tell it exactly what to change, what to keep, and what the final image should feel like. The model works best when the instruction is concrete and visual.

Keep refining the same image in steps

One of Nano Banana's practical advantages is multi-turn editing. Change the room first, then add objects, then adjust styling, then refine details. That lets you build the final image through controlled rounds instead of one risky prompt.

More AI Tools & Effects

Discover more tools and effects to power your creative workflow.

Nano Banana strengths that matter in real image editing

Native multimodal generation and editing

Nano Banana is built to work from both images and text instructions. That makes it useful for users who need generation and editing inside the same Gemini-native workflow instead of splitting the job across separate tools.

Conversational image refinement

Nano Banana supports everyday-language prompting and multi-turn editing. You can keep revising the same image step by step, which makes the workflow more practical for real editing tasks than one-shot prompt generation alone.

Likeness preservation for people, pets, and source images

Google introduced Nano Banana with a strong focus on keeping subjects recognizable during edits. That matters when a person, pet, or key visual element needs to survive scene changes, outfit changes, or repeated revision rounds.

Photo blending and style transfer workflows

Nano Banana is useful when multiple images need to become one scene or when the style of one image needs to influence another. That makes it practical for creative exploration, concept images, and edit-led visual storytelling.

Real-world knowledge inside the Gemini image stack

Because Nano Banana sits inside the Gemini image family, it benefits from the model line's stronger real-world knowledge and multimodal reasoning. That helps the output follow more believable scene logic and object relationships.

The original model in a broader Nano Banana family

Nano Banana is the original family base. Nano Banana Pro adds studio-quality precision and tighter creative control. Nano Banana 2 adds pro-level capability at Flash speed. Nano Banana remains the foundation for the series.

Nano Banana FAQ

These answers focus on the model itself, its Google DeepMind background, and how Nano Banana compares with Nano Banana Pro and Nano Banana 2.

Nano Banana is Google DeepMind's original Gemini image generation and editing model. It introduced the Nano Banana family with native multimodal creation and conversational image editing.