Google introduces Nano Banana 2 Lite and Gemini Omni Flash

Summary

• Google announced Nano Banana 2 Lite and Gemini Omni Flash
• Nano Banana 2 Lite generates text-to-image outputs in about 4 seconds
• Nano Banana 2 Lite is priced at $0.034 per 1K image
• Gemini Omni Flash supports video generation and conversational editing
• Gemini Omni Flash is in public preview and currently generates videos up to 10 seconds
• Both models are available in Google AI Studio, Gemini API and Gemini Enterprise Agent Platform

Google announced two new generative AI tools for image and video, Nano Banana 2 Lite and Gemini Omni Flash.

The company is now making Nano Banana 2 Lite available to developers as the fastest and most cost-efficient image model in the Gemini Image family, alongside Gemini Omni Flash, a model for video generation and conversational editing.

The announcement matters because it shows that Google wants to bring instant image generation, AI video creation and multimodal workflows closer together, giving creators and developers tools that can be combined into a single content production pipeline.

Nano Banana 2 Lite: speed and lower cost for image generation

Nano Banana 2 Lite, technically named gemini-3.1-flash-lite-image, is designed for fast image generation, early-stage ideation and applications that require high-volume output.

According to Google, the model can generate text-to-image outputs in about 4 seconds, making it useful for interactive prototyping, quick drafts and applications where speed is a critical factor.

The company also lists a price of $0.034 per 1K image, positioning Nano Banana 2 Lite as an option for developers who want to keep operating costs low without abandoning core quality features.

What changes compared to previous Nano Banana models

Google presents Nano Banana 2 Lite as the recommended upgrade for developers currently using the original Nano Banana, meaning gemini-2.5-flash-image.

The new model focuses on speed, lower cost and large-scale use, while, according to the company, retaining reliable prompt adherence, character consistency and better in-image text rendering.

The same family also includes other models. Nano Banana 2 works as a more balanced option for quality and cost, while Nano Banana Pro is aimed at more complex and professional use cases where accuracy matters more than speed.

Availability across Google apps and tools

Nano Banana 2 Lite is available in Google AI Studio, the Gemini API and Gemini Enterprise Agent Platform.

At the same time, Google is bringing it to consumer products, including AI Mode in Search, the Gemini app, NotebookLM, Google Photos, Stitch, Google Flow and Google Ads.

This shows that the company does not see the model only as a developer tool, but also as a technology that will gradually appear across many everyday Google applications.

Gemini Omni Flash: AI video with conversational editing

Gemini Omni Flash, technically named gemini-omni-flash-preview, is now available to developers through the Gemini API and Google AI Studio.

The model supports video generation and natural-language editing, using combinations of text, images and video as inputs. Google describes it as a high-quality and cost-efficient model for video generation and conversational editing.

The price mentioned by the company is $0.10 per second of generated video, the same as Veo 3.1 Fast.

What Omni Flash can do

Gemini Omni Flash is designed for workflows where the user can request changes to a video in natural language, without relying only on traditional editing tools.

The model can combine images, text and video as references in order to maintain greater control over the scene and consistency of the result. Google also says it uses Gemini’s knowledge to create more coherent videos, with better logic in action, narrative and content.

Another point highlighted by the company is text and action synchronization, allowing graphics or text inside the video to connect more directly with what is happening in the scene.

The limitations of Gemini Omni Flash

Despite the announcement, Gemini Omni Flash remains in public preview and has limitations.

For now, it supports video generations of up to 10 seconds, with Google saying that longer durations will come later.

Uploading audio references and scene extension are not yet supported in the Gemini API for this model. In addition, although the API schema accepts video references up to 3 seconds long, Google says they are not processed correctly by the model at this stage.

There are also limitations in character consistency when scenes change or when camera movements such as panning are used.

Combining image and video in one workflow

The most interesting part of the announcement is how Google suggests using the two models together.

A developer can use Nano Banana 2 Lite to quickly generate an image and then provide it as a reference to Gemini Omni Flash, turning it into an animated video.

Google also mentions the use of the Interactions API for multi-turn experiences, where history and context are preserved, allowing up to three sequential edits.

Safety and transparency

The two models use Google’s SynthID watermarking, so AI content can be identified and verified.

The company says AI content verification can be done through the Gemini app, Gemini in Chrome or Search. Transparency becomes especially important as image and video generation tools become faster, cheaper and easier to use.

What we think

Google’s announcement shows that the next battle in generative AI will not be decided only by quality, but also by speed, cost and the connection between different media types. Nano Banana 2 Lite is interesting for large-scale image generation and fast prototyping, while Gemini Omni Flash shows how AI video editing can become more accessible through natural language. The limitations of Omni Flash show that the technology is not yet mature for every professional scenario, but the direction is clear.

Frequently asked questions

What is Nano Banana 2 Lite?

It is Google’s new fast and cost-efficient image model in the Gemini Image family, designed for fast image generation and high-volume use.

How fast is Nano Banana 2 Lite?

Google says it can generate text-to-image outputs in about 4 seconds.

What is Gemini Omni Flash?

It is a Google model for video generation and natural-language editing, using text, images and video as inputs.

What are the main limitations of Gemini Omni Flash?

For now, it generates videos up to 10 seconds long, does not yet support audio references and scene extension in the Gemini API, and has limitations in character consistency during scene changes.

Where are the new models available?

Nano Banana 2 Lite and Gemini Omni Flash are available in Google AI Studio, the Gemini API and Gemini Enterprise Agent Platform, while Nano Banana 2 Lite is also coming to Google consumer products.