Use Google Web Search grounding to generate images based on real-time information
No image yet
Add a prompt to generate your first image.
Google DeepMind's speed-focused Gemini image model for fast 1K generation, prompt-based editing, and high-volume creative iteration without waiting on heavier production models.
Create clear app visuals, social graphics, product concepts, and thumbnails when speed matters more than maximum resolution.

Generate hero drafts, stickers, mood-board elements, and background visuals inside interactive design tools for faster creative loops.

Turn user photos or character references into location-themed postcards and lifestyle images for travel and entertainment apps.

Produce scene cards, character moments, and content previews quickly enough for interactive product experiences.

Draft product lifestyle scenes and campaign variations at scale, then promote stronger concepts to higher-resolution models when needed.

Generate usable 1K images in about four seconds for drafts, thumbnails, app visuals, and rapid prompt testing.
A lightweight option for frequent image requests, marketing variations, and in-app generation where every call does not need 2K or 4K.
Refine existing visuals with natural language—adjust objects, swap backgrounds, update styles, and polish compositions.
Keeps Nano Banana family strengths: solid prompt adherence, character consistency, readable in-image text, and practical real-world detail.
See credit costs for Nano Banana 2 Lite and every other model on the pricing page. Credits never expire.
View all pricingIt is Google DeepMind's speed-focused Gemini image model (also known as Gemini 3.1 Flash-Lite Image) for fast 1K generation, prompt-based editing, and high-volume creative iteration.
Nano Banana 2 Lite prioritizes lower latency and efficiency at 1K, while Nano Banana 2 offers stronger overall quality and higher-resolution options (1K/2K/4K) when polish and resolution matter more.
Nano Banana 2 Lite is optimized for 1K image output—ideal for drafts, thumbnails, app visuals, and frequent creative generation.
Yes. You can upload reference images and apply natural-language instructions to adjust objects, backgrounds, styles, and composition.
Text-to-image generation is positioned around the four-second range for 1K images, making it suitable for interactive tools and rapid prompt testing.
It supports a wide set of ratios including 1:1, 16:9, 9:16, 4:3, 3:4, ultrawide options, and auto.
Use Lite for fast drafts, frequent variations, and in-app generation. Switch to Nano Banana 2 or Pro when you need higher resolution, more complex production control, or maximum visual polish.
New users can usually start with free credits, then continue with flexible plans based on usage.