Google has rolled out two new generative-media models aimed squarely at developers who need speed and scale: Nano Banana 2 Lite, its fastest and most cost-efficient image-generation model yet, and Gemini Omni Flash, a high-quality model for video generation and conversational editing. Both were announced on June 30, 2026, by Google DeepMind and are available to developers immediately.

The release extends a wave of recent AI industry coverage focused on making generative media practical for production pipelines rather than just demos. Google is positioning the pair as tools that let builders connect rapid image generation with video creation in a single end-to-end workflow.

Nano Banana 2 Lite: Built for Speed

Nano Banana 2 Lite, identified by the model name `gemini-3.1-flash-lite-image`, is designed for rapid ideation and high-velocity developer pipelines where speed and cost are the primary constraints. According to Google, it delivers text-to-image outputs in 4 seconds, making it ideal for interactive prototyping and rapid visual drafting.

On cost, the model is priced at $0.034 per 1,000 images, a figure Google highlights as a cost-efficient choice for developers focused on drafting, managing operational budgets, or handling low-bandwidth usage. Despite prioritizing speed, the company says Nano Banana 2 Lite retains reliable prompt adherence, strong character consistency, and legible in-image text rendering — a recurring pain point for earlier image models.

Google is recommending Nano Banana 2 Lite as a replacement for developers still using the first version of Nano Banana (`gemini-2.5-flash-image`), promising immediate benefits across latency, cost, and quality.

A Growing Banana Family

To clarify where each model fits, Google laid out the full Nano Banana lineup:

  • Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image): Built for speed, optimized for near-real-time, high-volume workflows where ultra-low latency is critical.
  • Nano Banana 2 (Gemini 3.1 Flash Image): The generalist workhorse, delivering high quality at lower latency for the best balance of performance and cost.
  • Nano Banana Pro (Gemini 3 Pro Image): Optimized for complex, professional use cases, offering the most robust control and advanced reasoning where accuracy matters most.

The tiered approach mirrors how Google has structured its text models, giving developers a clear upgrade path from cheap-and-fast to premium-and-precise.

Gemini Omni Flash Brings Video to Developers

Alongside the image model, Google brought Gemini Omni Flash to developers for the first time. Described as a high-quality, cost-efficient model for video generation and conversational editing, Omni Flash lets users create and refine video through multi-turn, conversational instructions rather than a single rigid prompt.

Google framed the combination as the key to building comprehensive multimedia experiences: generate thousands of images rapidly with Nano Banana 2 Lite, then move into multi-turn video sequences with Omni Flash. The two models are designed to work together so creative iteration flows from still images to motion without switching platforms.

Where to Access Them

Both models are available across Google's developer and enterprise surfaces from launch day. Nano Banana 2 Lite and Gemini Omni Flash can be accessed through Google AI Studio, the Gemini API, and the Gemini Enterprise Agent Platform. They are also rolling out across Google consumer products, including AI Mode in Search, the Gemini app, and Google Flow.

The broad availability is strategic. By placing the same models in both developer tools and consumer apps, Google ensures that improvements flow in both directions — developer feedback sharpens the consumer experience, and consumer scale hardens the models for production use.

The Generative-Media Price War

The launches arrive amid a fierce price war in generative media. As image and video models have matured, the competitive frontier has shifted from raw quality to cost-per-generation and latency. At $0.034 per 1,000 images and 4-second generation times, Nano Banana 2 Lite is a clear shot across the bow at rival image models, pushing the entire market toward faster, cheaper output.

Product managers Alisa Fortin and Anish Nangia of Google DeepMind, who announced the release, emphasized that building with generative media is fundamentally about creative iteration — and that iteration only works when each step is fast and affordable enough to repeat freely.

Stay Ahead of AI

Generative media is evolving faster than ever, with new models arriving weekly. Track every release, benchmark, and pricing shift at AI Buzz Wire, your home for the latest AI developments.

Read more AI news →