Nano Banana 2 Lite

Google⁠ has introduced two major upgrades to its generative AI stack: Nano Banana 2 Lite, its fastest and most cost-efficient image generation model, and Gemini Omni Flash, a new multimodal video generation model built for conversational editing. The launch is aimed at developers and enterprises building high-volume creative AI applications.

Nano Banana 2 Lite: Fast, Cheap, and Built for Scale

Nano Banana 2 Lite is the latest addition to Google’s Gemini Image model family and is optimized for speed, throughput, and low cost. Google says the model can generate text-to-image outputs in around 4 seconds, making it ideal for rapid iteration, prototyping, and large-scale image generation pipelines.

Key highlights include:

This makes the model particularly useful for:

Google is rolling it out through Google AI Studio, Gemini API, and Gemini Enterprise Agent Platform, with consumer availability expanding to Search AI Mode, Gemini App, Google Photos, and more.

Gemini Omni Flash Brings Conversational Video Editing

Alongside Nano Banana 2 Lite, Google also released Gemini Omni Flash in public preview. This model focuses on video generation and conversational editing, allowing users to create or modify videos using natural language instructions.

Unlike traditional text-to-video systems, Gemini Omni Flash supports multimodal inputs including:

Users can perform advanced edits such as:

Google says this gives creators much finer control over video production without restarting from scratch after every edit.

Image-to-Video Workflow

The most interesting part of the launch is how both models work together.

Developers can generate images quickly using Nano Banana 2 Lite and then feed those outputs into Gemini Omni Flash to animate them into videos. This creates a streamlined image-to-video AI pipeline for production-scale creative workflows.

This workflow could significantly reduce production costs for: