Google has invited Gemini Advanced subscribers to try out Veo 2, its AI-powered video generator that the company says can create high-definition cinematic videos based on text descriptions. Those who want to try out the algorithm can simply select it from the list of available tools, after which they can generate an 8-second video in 720p resolution based on a text prompt.
Image source: Google
Google’s proposal includes a limit on the number of videos created with Veo 2. The company did not provide more details, noting only that users will receive a corresponding notification when approaching the limit. It is also known that Veo 2 generates videos in MP4 format, and mobile users will be able to immediately upload the created content to TikTok or YouTube using the “Share” button.
According to Google, the updated AI model “better understands real-world physics and human motion,” allowing it to deliver “smooth character movement, realistic scenes, and finer visual detail across a range of stories and styles.” Videos generated by Veo 2 are accompanied by digital markers called SynthID, which indicate that the video was created using a neural network.
At the same time, Google is opening up Whisk Animate, a tool that uses Veo 2 to create 8-second video clips from an image, to One AI Premium subscribers. Whisk Animate is available to subscribers worldwide through Google Labs.