“From reimagining historical events through the eyes of a modern influencer to exploring what it would sound like to cut through a glass apple, to quite a few sightings of the mythical Bigfoot, your imagination is the limit when you create photos and videos on Gemini. These amazing creations are why we on the Veo 3 team do what we do — and why we’re making Veo 3 available to more people around the world,” said Google describing the video model’s capabilities.

Google revealed Veo 3, their latest advancement in AI video production, during its annual developer conference on May 20. In addition to producing visually attractive and cinematic videos, this model also incorporates realistic audio, such as spoken conversation, sound effects, and background music, making the videos seem real.

All the videos generated using Veo 3 from photos will have a visible watermark and are embedded with an invisible SynthID watermark. These additions to the output indicate that the contents are made using AI. 

Meanwhile, Google has also asserted its commitment to safe and responsible use of AI. The company said that it continues to take important steps to ensure video generation is safe. Google is ensuring this by the means of extensive red teaming and evaluations of its AI models.

After Veo 3 was introduced at Google I/O, several X users had shared their creations using the model. The AI tool was hailed as Google’s answer to OpenAI’s Sora. Following the teaser, Eli Collins, Google’s DeepMind product vice president, said that Veo 3 excels from text and image prompting to real-world physics and accurate lip syncing.