Google has unveiled a powerful new upgrade to its Gemini AI ecosystem-this time expanding beyond chat capabilities to include dynamic video creation. The tech giant is launching its state-of-the-art video generation model, Veo 2, via the Gemini mobile and web platforms, providing Gemini Advanced subscribers with access to AI-powered short video creation based on detailed text descriptions.

Veo 2 operates in much the same way as other next-gen video generation tools like OpenAI’s Sora. Users simply enter a comprehensive text prompt, and the AI processes that input to create animated clips. Google claims Veo 2 has been engineered with an advanced understanding of physical realism, particularly in simulating human motion in lifelike ways. The example videos released by the company highlight visually impressive results, although it’s worth noting that these samples likely showcase the model at its best.

Initially, Veo 2 will be accessible through the model selection menu in the Gemini interface, but Google has indicated that its exact placement could change as integration deepens and the feature evolves. As with past rollouts, access to Veo 2 will be phased in gradually, and it may take several weeks before the majority of Gemini Advanced users can start using it. This gradual release mirrors earlier feature deployments like Gemini Live, which also saw a staggered distribution post-announcement.

Once it becomes available, Veo 2 will support highly specific text inputs, giving users precise control over the kind of content the AI produces. For now, the model can generate videos lasting up to 8 seconds at 720p resolution. These clips are downloadable in the common MP4 format, making them easy to use across different devices and platforms.

Given the resource-intensive nature of video generation, Google has introduced a monthly usage cap for Veo 2. However, the company has yet to disclose the exact limit, only stating that users will be alerted as they approach their monthly quota.

Topics #AI #Artificial Intelligence #Gemini App #Google #Veo 2 #Video AI #Video Generation