“We are still a couple of years away from my seven-year-old scripting and creating her own full-fledged Pixar film”, wrote Joel Braithwaite head of agency partnerships Southeast Asia at Google in an article published late last year addressing the potential of generative AI.
Going by the videos created using Google Cloud’s state of the art video generation model Veo3.1, it is safe to assume that a seven-year-old creating an animated masterpiece could happen within that timeframe. The most recent iteration of Veo3.1 comes with several improvements and is a testament to the rapid advances in generative AI.
While it took a little over three decades for the motion picture industry to move past the silent era, and six decades to shift completely from black and white to colour; AI-powered video generation has upgraded from blocky, uncanny, low-resolution visuals to cinematic films in just three years. Veo 3.1 is available to creators and prosumers through tools like Flow and is available for businesses and developers to build with at scale via the Google Cloud Vertex AI platform. . Launched a few months ago, Flow has spurred a creative renaissance of sorts with over 275 million videos, counting those powered by earlier iterations of Veo.
However, Veo3.1 is not just helping people with no previous filmmaking experience (not to mention no equipment, or budget) tap into their inner Scorsese, Cameron, or Nolan. Among its foremost use cases is in advertising—an industry that is well acquainted with the value of a resonant video in building interest, affinity, and loyalty.
What Veo3.1 helps agencies and marketers do better than any of its predecessors is effortlessly repurpose brand assets—preexisting as well as new—into video.
Making video easier than ever before
With Ingredients to Video (I2V), a reference image, or a series of images—for example, a character, a location, or objects—can swiftly be fused to create a fully formed scene, complete with sound. The visual language of the final film will reflect the aesthetic of the input—be it a picture or a video. Reference images can also help maintain the consistency of a character across multiple films.
‘Extend’ allows the last few frames from a pre-existing clip to be used to create a completely new and longer scene in the style of the original film. Enhanced camera controls allow greater freedom in the framing and exact movement of shots.
Objects within a video can be added or removed at will. While making these modifications, Veo considers scale, interactions between artifacts in the video, and lighting. Veo also allows users to define the path of various objects within a film.
Frames to Video (F2V) ensures that a film can be created with just the start and end visual. Veo3.1 generates an interstitial video, doing all the heavy lifting of building a convincing narrative. With the Outpainting feature, an existing video can be transformed and expanded into a desired screen size or shape.
Naturally, these videos can be improved upon and finessed with further prompt engineering.
With these changes under the hood, Veo3.1 demonstrates benchmark leading performance when pitted against other video generation models, in head-to-head comparisons by human raters. The improvements are in enhanced realism, stronger prompt adherence and improved audiovisual quality, capturing true-to-life textures.
These improvements are vital to an era in which video is ubiquitous: consumed in a variety of formats—from short clips to long form—and where the orientations range from horizontal to vertical. The importance of reaching audiences irrespective of their video consumption preference with a relevant mix of formats has already been well-established; Veo3.1 makes creating, crafting, or reworking assets to meet this goal easier than ever before. At a time when the ROI of every single asset is under severe scrutiny, AI-powered video generation enables the rapid creation of new assets or extracting even greater mileage out of archival footage. Apart from its positive impact on timelines and budgets, the greatest advantage is being able to tap into possibilities around creativity, personalisation, and relevance that would have been inconceivable even half a decade ago.
For insider insights into unleashing the power of Veo3.1 on Google Cloud’s Vertex AI platform, be sure to tune in to our upcoming webinars on November 20, 2025, and December 5, 2025. Scheduled to be held between 11:30 am and 12:30 pm SGT, the webinars will include perspectives from industry experts and practitioners aimed specifically at brand creators and managers. With ample time set aside for a Q&A session, even those who have been merely dabbling with AI so far can quickly level up and embark on the journey to become an expert user.
Watch out for more details on speakers, and the agenda over the next few weeks here.