Google says Veo, a rival to OpenAI’s Sora, will help marketers and advertisers produce high-quality, AI-generated content for campaigns.
Google on Tuesday announced it is making its most advanced generative AI video tool, Veo, available to Cloud customers using its Vertex AI platform. With the rollout, Google hopes to seize an edge over rivals like OpenAI’s anticipated video generator, Sora, as AI companies race to attract advertisers, marketers and brands and transform the industry.
Veo is capable of generating “high-quality, high-definition” 1080p resolution videos from text or image prompts in a range of cinematic and visual styles. The tool can produce realistic clips of animals, people and objects “with exceptional speed,” said Vertex senior director of product management Warren Barkley in a blog post announcing the launch, adding that the model has an “advanced understanding of natural language and visual semantics” to ensure generated video closely matches prompts.
The tool, first unveiled at Google’s I/O developer conference earlier this year, will be available in private preview for businesses using Vertex AI. Google did not state in the preview release when or if it would make the model widely available nor did it outline restrictions on how long generated clips can be. Veo’s section on Google DeepMind’s website, the tech giant’s artificial general intelligence lab behind the video model, loosely suggests clips can be generated or extended to a minute “and beyond.”
Google’s ‘most capable video generation model’
Barkley describes Veo as “our most capable video generation model” and says the tool can help brands tap into new levels of creativity, streamline video production workflows and “transform their existing creative assets into dynamic visuals.”
The blog shows examples of text and image prompts used to generate video with Veo. This includes creating short clips based on AI-generated images of a man reading and of a dog, real-life images of marshmallows roasting over a fire and a crowd of people at a concert and text-based prompts of a sailing boat at sea and a teddy bear wearing shades playing an electric guitar in front of a waterfall.
The brands and agencies tapping genAI for campaigns
Google also said it will release its “highest quality image generation model,” Imagen 3, to all Vertex AI customers from next week. The Imagen 3 model can craft realistic and high-quality images from simple text prompts in an array of visual styles, including oil paintings, comic book illustrations and photography. Barkley said businesses can use Imagen 3 to create “images that reflect their own brand style and logos for use in marketing, advertising, or product design.”
Google said its image and video generating AI tools have already been tapped by brands like food giant Mondelez International, ad agency WPP and digital travel platform Agoda to boost and streamline content pipelines.
“With Veo and Imagen, we are narrowing the gap between imagination and execution, enabling our people to develop high-quality, photo-realistic, campaign-ready visuals in a matter of minutes,” said WPP chief technology officer Stephan Pretorius.
“Our collaboration with Google Cloud has been instrumental in harnessing the power of generative AI, notably through Imagen 3, to revolutionize content production,” said Jon Halvorson, SVP of Consumer Experience & Digital Commerce at Mondelez. “This technology has enabled us to produce hundreds of thousands of customized assets, enhancing creative quality while significantly reducing both time to market and costs,” Halvorson said, adding that Veo will help expand this to video, “further streamlining production processes and setting new benchmarks in marketing.”
Beating Sora
Veo represents Google’s response to leading AI video generators like OpenAI’s Sora and Runway’s Gen-3. It comes as tech companies jostle to tap into the growing market of businesses eager to use generative technology to improve processes and streamline performance. Google claims 86% of enterprise companies using generative AI in production report an increase in revenue.
Veo’s launch, albeit in private preview, adds pressure to OpenAI over the whereabouts of Sora. The Microsoft-backed company shut down early access to the hotly anticipated video generator model after a group of testers leaked a demo version in November and published an open letter savaging OpenAI for allegedly exploiting artists for “art washing” and unpaid labour. In a statement, OpenAI stressed participation in the research preview was voluntary, with no obligation to provide feedback or use the tool.