Google has announced the availability of its new Veo 2 model for video generation through the Gemini Advanced service, available to paid subscribers. This move appears to be part of Google’s broader strategy to keep up with growing competition from platforms like OpenAI’s Sora as well as Runway's newest model Gen-4.
Veo-2 allows users to create 8-second videos in 720p resolution, with a built-in sharing button that enables direct posting to platforms like TikTok and YouTube through the Gemini app.
Veo 2’s updated version brings improved understanding of object motion and realistic physics, helping generate cinematic scenes with greater visual detail. Users can describe scenes in detail to transform written ideas into videos, choosing from a range of styles that span from realism to imagination.
In addition, Google has linked Veo 2 with the Whisk tool, currently an experimental feature under Google Labs. Whisk uses user-uploaded images combined with text prompts as a starting point to create new visuals.
This connection has resulted in a new tool called "Whisk Animate," which launched in March as a research preview within Google Labs for a limited group of users.
Whisk Animate is now available exclusively to subscribers of the Google One AI Premium plan. It transforms images created via Whisk into short, 8-second video clips, generated by Veo 2.
As part of its policy, Google has set standards for downloading generated videos, including the addition of a digital watermark (SynthID).
The company also mentioned a monthly limit on the number of videos users can create with Veo 2.
At present, the service is not available for educational or enterprise Workspace accounts.
Google seems clearly focused on integrating its video generation technology across multiple projects. One example is its February rollout of Veo 2 in YouTube Shorts, which allowed users to create short-form content using the model.
Looking ahead, Google announced that upcoming updates will integrate Veo 2 with Gemini models to deepen their understanding of the physical world — a step that could open new possibilities in the field of generative AI.
