Google has unveiled the latest iteration of its video generation technology, Veo 2, which is now available for Gemini Advanced users and Whisk experimental testers. This innovative model is designed to transform user-generated text descriptions into engaging eight-second video clips, perfect for sharing on platforms like YouTube Shorts and TikTok.
With Veo 2 integrated into Gemini, users can craft a variety of video content, from short narratives to specific visual concepts. The effectiveness of this feature hinges on the detail provided in the text prompts, encouraging users to be as descriptive as possible to achieve the desired results. A new “Share” button has been introduced for mobile users, streamlining the process of uploading their creations to social media.
As Google rolls out these enhancements globally, it emphasizes its commitment to safety and compliance, ensuring that all generated content adheres to its policies. Each video produced with Veo 2 will be tagged with SynthID, clearly indicating that it is AI-generated. Additionally, Gemini Advanced users will encounter a monthly limit on the number of videos they can create, with notifications to alert them as they approach this threshold.
Getting into Whisk
The Veo 2 model also makes its debut in Google’s experimental platform, Whisk, introducing a feature called “Whisk Animate.” This allows users to animate their generated images into cinematic-quality video clips, breathing life into their creative visions. Users can expect vibrant and dynamic visuals in these eight-second animations.
Whisk Animate is being rolled out today, but access is limited to subscribers of the Google One AI Premium plan. This follows the previous updates announced in December, which included enhancements for both Veo 2 and Imagen 3, along with the launch of Whisk itself. The platform encourages users to “Prompt less, Play More,” allowing them to generate content by simply providing images that meet three criteria: subject, scene, and style.
While Whisk aims to simplify the creative process, Google acknowledges that the AI may occasionally misinterpret key details, such as hair or skin tone. To address this, Whisk includes a “Review & Edit” feature, empowering users to refine the AI’s output to better align with their expectations.
