Google has introduced its new and improved Veo 2 AI-powered video generator, which it claims beat out rivals like Sora and offers higher quality realism, better fidelity, advanced motion capabilities, and greater camera control options.

“Veo 2 outperforms other leading video generation models, based on human evaluations of its performance,” said Google DeepMind in a blog post.

Users could sign up to try Veo 2 via VideoFX. Google’s launch of Veo 2 comes days after competitor OpenAI launched Sora Turbo and allowed general users to access its text-to-video generation model months after it was first announced.

Google shared a range of video clips that showed Veo 2 was capable of generating hyper-realistic scenes featuring animals and people, apart from animated clips. While the motions were not entirely accurate in places, many of the scenes had consistent detailing and did not violate the laws of physics through morphed objects, distorted bodily movements, etc. at first glance. The lighting and framing were also engaging.

However, Google DeepMind acknowledged there was still work to be done.

“While Veo 2 demonstrates incredible progress, creating realistic, dynamic, or intricate videos, and maintaining complete consistency throughout complex scenes or those with complex motion, remains a challenge. We’ll continue to develop and refine performance in these areas,” said Google DeepMind.