Google has introduced Nano Banana 2 Lite, the latest version of its in-house artificial intelligence model for image generation, offering significantly faster performance and lower operating costs than previous releases.
The company said the new model can generate images in approximately four seconds, making it suitable for high-volume creative workflows that require rapid iteration.
Faster Image Generation at Lower Cost
According to Google, Nano Banana 2 Lite has lower latency than earlier versions and is priced at US$0.034 per 1,000 generated images, positioning it as a cost-effective option for developers and businesses creating large volumes of visual content.
The release follows the debut of the original Nano Banana last summer, powered by Gemini 3.1 Flash, and the launch of Nano Banana 2 in February, which introduced improved image realism. Google also continues to offer Nano Banana Pro for more advanced image generation use cases.
The company described Nano Banana 2 as its general-purpose image generation model, while Nano Banana 2 Lite is optimized for speed and large-scale production.
Expanding Google’s Generative Media Platform
Nano Banana 2 Lite is now available through Google AI Studio, the Gemini API and Google’s Gemini Enterprise Agent Platform. The new model replaces the original Nano Banana, which Google now classifies as its legacy image generation model.
Google also announced the broader availability of Gemini Omni Flash, its AI-powered video generation model first introduced at Google I/O earlier this year. The company said Omni Flash is priced at US$0.10 per second of generated video.
In addition, Google demonstrated Omni Product Studio, a new application designed to transform static AI-generated images into cinematic e-commerce videos.
Focus on Creative Workflows
Google said the latest releases are intended to help developers build end-to-end multimedia experiences by combining rapid image generation with AI-powered video creation and editing.
“Building with generative media is often about creative iteration.”
The company added:
“With these two models, developers can build comprehensive, end-to-end multimedia experiences that connect rapid image generation with video creation and editing.”