Description
Nano Banana is the name Google DeepMind uses for its Gemini Image family of models, built for generating and editing images from text prompts and uploaded photos. The models use Gemini's language understanding to read a prompt closely, so the result lines up with a fairly specific request rather than a generic interpretation of it.
Read More
Three capabilities run through the family. Multimodal understanding lets a user upload an image and give text instructions to create complex, detailed edits or new images from that starting point. Conversational input means prompts can be written in everyday language, and a user can keep refining an image across a back-and-forth exchange rather than writing one perfect prompt. Real-world knowledge, drawn from Gemini's reasoning, means generated images tend to follow real-world logic in lighting, physics and object relationships rather than looking obviously synthetic.
Google DeepMind offers Nano Banana across a few tiers built on different Gemini versions. Nano Banana Pro, built on Gemini 3, targets studio-quality precision and control for detailed edits. Nano Banana 2, built on Gemini 3.1 Flash Image, aims for pro-level generation and editing at faster Flash-tier speed. Nano Banana 2 Lite, on Gemini 3.1 Flash-Lite Image, is positioned as the fastest and lowest-cost option in the family for high-volume generation and editing.
The models are accessible through several Google surfaces: directly inside the Gemini app and Google Pics for conversational image creation, through Google AI Studio for building with the models, and via the Gemini API for developers who want to integrate image generation and editing into their own products. DeepMind also publishes a prompt guide covering how to describe subjects, settings and mood for more predictable results.
Nano Banana suits designers, marketers, developers and anyone who wants to generate or edit images by describing what they want rather than using manual editing tools, from quick playful images to more exacting, detailed edits on the Pro tier.
- CategoryAI Models & LLMs
- Pricing ModelCost-effective
- DeploymentCloud
- PlatformsWeb
Nano Banana Highlights
Text-to-image generation from natural-language prompts
Image editing from an uploaded photo plus text instructions
Multimodal understanding that reads text and image input together
Conversational refinement across multiple turns instead of a single prompt
Real-world logic in generated lighting, physics and object placement
Tiered model family: Nano Banana Pro, Nano Banana 2 and Nano Banana 2 Lite
Access through the Gemini app, Google Pics, Google AI Studio and the Gemini API
Published prompt guide for more predictable results
Getting Started with Nano Banana
Pick a way in: Open the Gemini app or Google Pics for quick use, or Google AI Studio or the Gemini API to build with the model.
Write a prompt: Describe the subject, setting and mood you want, or upload a photo to edit.
Add instructions: Combine the image with text instructions for the specific edit or style.
Refine conversationally: Ask for changes in plain language and keep adjusting the result.
Choose a tier: Use Nano Banana Pro for precise, studio-quality edits or Nano Banana 2 / 2 Lite for faster, lower-cost generation.
Nano Banana's Use Cases
- Marketing image creation
- Photo editing by description
- App and product prototyping
- Concept and mood exploration
