Google announced two generative artificial intelligence models on June 30, 2026, targeting faster image generation and video editing workflows. The company released Nano Banana 2 Lite and expanded availability for Gemini Omni Flash across its developer and consumer platforms.
Nano Banana 2 Lite, officially designated gemini-3.1-flash-lite-image, is Google’s fastest image model to date. The model delivers text-to-image outputs in 4 seconds at a price of $0.034 per 1,000-resolution image. Google positioned it as a direct replacement for its legacy gemini-2.5-flash-image model, offering higher throughput for developers building high-volume drafting pipelines.
The new image model is available immediately in Google AI Studio, the Gemini API, and the Gemini Enterprise Agent Platform. Google is also rolling the model out across consumer surfaces, including AI Mode in Search, the Gemini app, NotebookLM, Google Photos, Stitch, Google Flow, and Google Ads. It sits alongside the standard Nano Banana 2 generalist model and the higher-capability Nano Banana Pro.
Video Generation and Editing
Alongside the image model, Google launched Gemini Omni Flash in public preview for developers via the Gemini API and Google AI Studio. The model, designated gemini-omni-flash-preview, handles video generation and conversational editing using prompts that combine text, images, and video. Google set the price at $0.10 per second of video output, matching its Veo 3.1 Fast model.
Omni Flash is also accessible in the consumer Gemini app and Google Flow. Google noted several technical limitations for the preview release. The model currently restricts video generation to 10 seconds and cannot process audio reference uploads or scene extensions in the Gemini API. While the API schema accepts video references up to 3 seconds, the model cannot correctly process them yet.
Workflow Integration
Google designed both models to operate together in multi-turn developer workflows. Under this setup, Nano Banana 2 Lite generates a base image that Gemini Omni Flash subsequently animates into video. Using Google's Interactions API, applications can maintain session history across up to three sequential edits. Both models embed SynthID watermarking into outputs to allow content verification through Search, Chrome, and the Gemini app.
