Google’s new AI video generation rivals OpenAI’s Sora

Google Deepmind has introduced Veo 2, a new artificial intelligence (AI) video generation tool that builds on the original Veo and creates “incredibly high-quality videos” in a move to beat OpenAI at their own game.

The next iteration of Google’s flagship text-to-video tool Veo, Veo 2 can create “minutes in length clips” in 4k resolutions and Google has emphasized its understanding of cinematographic requests, stating in their press release: “Suggest “18mm lens” in your prompt and Veo 2 knows to craft the wide angle shot that this lens is known for, or blur out the background and focus on your subject by putting “shallow depth of field” in your prompt.”

According to Google, Veo 2 is less likely to “‘hallucinate’ unwanted details”, and has an “improved understanding of real-world physics and nuances of human movement and expression”.

Today, we’re announcing Veo 2: our state-of-the-art video generation model which produces realistic, high-quality clips from text or image prompts.