Creating an image with AI is no longer astonishing; the real challenge lies in the ability to modify, extend, and transform an idea into a more intricate narrative without losing coherence. When it comes to video generation, the complexity multiplies due to the factors of movement, time, and character consistency. Gemini Omni emerges as a groundbreaking solution to these challenges, promising to simplify the editing process.
Google DeepMind positions Gemini Omni as a video tool akin to Nano Banana, its picture-generating counterpart that revolutionized visual creation. Launched in August 2025, Nano Banana rapidly gained traction, acquiring 13 million users within four days and generating over 5 billion images by mid-October.
Introducing Gemini Omni Flash
Gemini Omni Flash serves as the inaugural model in the Gemini Omni line. Designed to create content from any entry, the platform allows users to blend images, audio, video, and text to generate high-quality videos enriched by Gemini’s comprehensive real-world knowledge.
A Model Committed to Coherence
One of the most captivating aspects of Gemini Omni is its editing process. Rather than merely generating clips from scratch, it enables users to modify existing scenes with a sequence of instructions. Users can fine-tune elements such as aesthetics, action, environment, camera angles, and styles while ensuring character consistency and scene cohesion.
For instance, Gemini Omni can transform a scene based on direct instructions like changing an object’s material or altering an action. Below are some illustrative prompts:
- “Make the sculpture out of bubbles.”
- “When the person touches the mirror, make the mirror ripple beautifully like liquid, and the person’s arm turns into reflective mirror material.”
- “Claymation explainer of protein folding, everything is made out of clay, no hands, stop motion, accurate.”
In an initial test conducted by Xataka, a static photograph of the Puerta de Alcalá in Madrid served as the starting point. The prompt was simple:
- “Create a video from this image. Cars are moving forward, and people are walking.”
The resulting video showcased the animated original image, featuring moving cars and pedestrians accompanied by fitting ambient sounds. While some logos, like that of Mercedes-Benz, were discernible, others, such as Fiat, appeared less so.
Availability and Limitations
As for accessibility, Gemini Omni Flash is rolling out to Google AI Plus, Pro, and Ultra subscribers through Gemini and Google Flow. Additionally, users can expect its free release in YouTube Shorts and the YouTube Create App soon.
However, in testing with a corporate account, limitations became evident. After generating three videos, a notification stated that the video generation limit had been reached until a specified date. This constraint is likely due to the resource-intensive nature of AI video creation, indicating that Google may be moderating access during this initial phase.
Looking Towards the Future
When discussing AI video generation, names like Sora often come to mind. While Sora was initially seen as a leading contender in the field, its trajectory has been noticeably shorter, with its website and app becoming unavailable by April 2026, though its API will continue operating until September 24.
As the landscape of AI video generation evolves, Gemini Omni could set a new standard for creativity and coherence in video storytelling, much like its predecessor Nano Banana did for images.