Introduction
In a strategic move aimed at strengthening its position in the advanced technology market, tech giant Google announced on Tuesday the launch of two new and innovative multimedia models entirely powered by artificial intelligence. This announcement represents a major and notable expansion in the company’s suite of generative AI tools, designed specifically to make digital image and video creation much faster and significantly less expensive, yielding great and direct benefits for technical developers, content creators, and everyday consumers alike. This technological development reflects the company’s ongoing commitment to delivering advanced solutions that meet the growing needs of the fast-paced digital content industry and provide flexible tools serving various creative directions worldwide.
- Blazing speed and extremely low cost
- Innovation in advanced video generation
- Broad availability of models across Google products
- Major industry partnerships and content credentials
- Frequently asked questions
Blazing speed and extremely low cost
The “Nano Banana 2 Lite” model, which is now generally available to all users, is the fastest image generation model Google has released in its technical history. This innovative model features exceptional capabilities that enable it to produce high-quality visual images in a very short time of no more than 4 seconds. Alongside this blazing speed, the model comes with an extremely low operating cost of just $0.034 USD per image produced at a resolution up to 1,000 pixels, known as 1K resolution. From a purely technical standpoint, this model bears the official designation “Gemini 3.1 Flash Lite Image” and was primarily designed and developed as an upgraded, more efficient alternative to the original model known as “Nano Banana.” This new release offers notable and fundamental improvements in visual quality, the level of character consistency across multiple images, and superior accuracy in rendering text written inside images. This model has been engineered and built specifically to handle use cases requiring high production volume and complex operations, such as A/B testing multiple commercial ad variants to select the best ones, as well as running interactive social applications on a very large scale that suits the demands of a crowded and fast-paced global market.
Innovation in advanced video generation
In the same technical context, the “Gemini Omni Flash” model entered public preview on the same day the previous model was released. It is a smart model that was first previewed and presented during Google’s developer conference held last May. This highly advanced model allows users to create and edit videos interactively and continuously using simple and direct natural language prompts. The power and flexibility of this model stand out in its unique ability to combine diverse and complex inputs—including written text, still images, and pre-existing video clips—within a single workspace to produce innovative cinematic output. The pricing for using this model is $0.10 USD per second of video output. Currently, the model produces short video clips up to a maximum duration of 10 seconds, with confirmed and announced future plans to expand this timeframe to include much longer clips. Furthermore, the model features the remarkable ability to generate native audio files and effects integrated alongside the produced video clips, while maintaining complete consistency of character style and visual scenes across all edits and modifications made by the user during interactive and ongoing conversations with the AI tool, thereby reducing the need for complex and expensive editing software.
Broad availability of models across Google products
To ensure these innovative technologies reach the widest possible audience of beneficiaries, Google ensured that both models are available for use through multiple digital channels. Professional developers can easily access these models via Google AI Studio, the Gemini model API, and the Gemini Enterprise Agent platform. As for individual users and consumers worldwide, the company has begun rolling out these models gradually and thoughtfully through the official Gemini mobile app, the AI-powered search engine experience, the NotebookLM smart notes app, Google Photos, the Google Flow tool, and the widely used Google Ads platform.
Among the practical and prominent applications of these technologies is NotebookLM, which currently utilizes the capabilities of the Nano Banana 2 Lite model to power a new and innovative feature called “Short Video Overviews.” This smart feature condenses and summarizes long documents and text files uploaded to the platform, turning them within seconds into vertical video clips lasting about 60 seconds. These innovative clips include clear voice narrative explanations and engaging educational animations to facilitate content comprehension for modern readers and researchers. This educational and practical feature is currently being rolled out exclusively to English-speaking users aged 18 and older across web browser platforms as well as Apple’s Android and iOS mobile operating systems, providing an exceptional and easily digestible knowledge experience.
Major industry partnerships and content credentials
The scope of the launch was not limited to the company’s own products and services; Google also proudly highlighted a group of early industry partners who quickly moved to adopt these advanced technologies for integration into their own services and applications. This distinguished list included prominent global companies such as Adobe, a leader in graphic design software, major advertising group WPP, and modern collaborative design platform Figma. In this regard, Matt Chوتين, senior product manager at Adobe, made an important statement confirming that these advanced and fast models will be seamlessly and deeply integrated into the popular Adobe Firefly platform. He explained that the goal of this integration is “to help creators and digital artists move faster and more efficiently from a preliminary idea to final visual content ready for publication and commercial use across various fields.”
To ensure the highest levels of safety, reliability, and transparency in the modern digital era where fake news proliferates, all the new models mentioned feature default, built-in invisible SynthID watermarks, alongside content credentials based on the globally recognized C2PA standard. This effective step helps track the origin of AI-generated content and verify its authenticity, preventing tampering and strongly helping to avoid misinformation and document the true sources of images and videos circulating on the internet.
Frequently asked questions
Question: What is the operating cost of Google’s new image generation model?
Answer: The new model provides the ability to generate high-resolution images at an extremely low cost of just $0.034 USD per image reaching up to 1,000 pixels in resolution.
Question: What is the maximum duration for video clips currently generated by the new model?
Answer: The model currently produces video clips up to a maximum duration of 10 seconds, with confirmation of future plans to significantly expand this timeframe.
Question: How will the smart notes application benefit from these models?
Answer: The application uses the new model to transform text documents uploaded by users into 60-second educational vertical video clips equipped with explanations and animations.
Question: What content protection and transparency technologies are adopted in these models?
Answer: The models rely on integrating hidden SynthID watermarks and standard content credentials by default to track the media source and ensure its reliability.
