On June 30th (local time), Google announced two new AI models: “Nano Banana 2 Lite,” the latest product in the Nano Banana image generation series, and “Gemini Omni Flash,” a new model designed for video generation and editing. Both were made readily available to developers and represented a major push to accelerate creative production while reducing costs.
Nano Banana 2 Lite is positioned as the fastest and most cost-effective product in Google’s lineup of image generation models. The company claims it can generate images from text in just 4 seconds, and has aggressive pricing of $0.034 per image at 1K resolution. Google describes this model as ideal for rapid ideation and high-frequency workflows that require continuous generation of large numbers of images. The technical name is “Gemini 3.1 Flash Lite Image” and it is a lightweight version of the general-purpose model “Nano Banana 2” released in February.
With this launch, Google’s image generation portfolio now has a three-layer structure: “Nano Banana Pro,” “Nano Banana 2,” and “Nano Banana 2 Lite.” The top-of-the-line Nano Banana Pro targets complex and specialized use cases. The mid-range Nano Banana 2 serves as a general purpose core model. The newly introduced Nano Banana 2 Lite strives for maximum speed and minimum cost. The original Nano Banana is currently classified as a legacy model, and Google recommends users to migrate to Nano Banana 2 Lite.
Nano Banana 2 Lite can be accessed through Google AI Studio, Gemini API, and Gemini Enterprise Agent Platform. It will also gradually roll out to Google’s consumer products, including AI mode in search, the Gemini app, NotebookLM, Google Photos, and Google Ads. Images generated by the model are embedded with SynthID, a digital watermarking technology that identifies AI-generated content.
Meanwhile, Gemini Omni Flash, which was launched for developers on the same day, is a high-quality, low-cost video generation model built on the “Gemini Omni” platform announced at Google I/O 2026. Its distinguishing feature is its ability to generate videos from combined text, image, and video inputs as well as support conversational video editing using natural language. Pricing is set at $0.10 per second of video, matching the pricing for Google’s existing video generation model Veo 3.1 Fast. We currently support generating videos up to 10 seconds in length, but we plan to support longer videos in the future. It can also be used from the Gemini app and the video creation tool “Google Flow”.
Google explained the differences between the two models, noting that Veo 3.1 excels at producing high-quality one-off clips, while Gemini Omni Flash focuses on interactive, iterative editing that combines multiple assets, such as images and videos. The company seems to be positioning the two models for different use cases.
Google also introduced specific ways to combine Nano Banana 2 Lite and Gemini Omni Flash. For example, images generated by the former can be passed to the latter to create a video, or the “interaction API” can be used to perform up to three consecutive editing rounds while preserving the session history. Demo applications that combine both models were also released, including “Anywhere,” which combines users’ photos with famous tourist spots around the world to create a video, and “Space Lift,” which allows users to generate interior design concepts from room photos and review them in video.
Alongside these announcements, Google revealed that personalized image generation powered by Nano Banana in the Gemini app is now available for free to all users in the United States. Previously, this feature was limited to paid subscribers on Plus, Pro, and Ultra plans, but this rollout significantly expands access. This feature automatically generates images that reflect your personal preferences based on information you give us access to from Gmail, Google Photos, YouTube, and Google Search. Users can obtain their favorite images by simply inputting “an illustration depicting themselves or their hobbies” without having to write detailed instructions.
According to TechCrunch, Google recently signed a massive $75 million deal with popular indie studio A24, but the partnership has drawn heavy criticism from some fans. Corporate investment continues to accelerate despite persistent concerns that AI-generated images and videos will fuel a flood of low-quality content, dubbed “AI slop.” Gemini has already surpassed 750 million monthly active users this year, demonstrating Google’s increasingly solid presence in the AI market.
Google says on its official blog, “Developing with generative media often revolves around creative iteration. These two models enable developers to build comprehensive, end-to-end multimedia experiences that marry rapid image generation with video creation and editing.”
