Article originally posted on InfoQ. Visit InfoQ

Google Cloud has introduced Veo and Imagen 3, two new generative AI models available on its Vertex AI platform. Veo generates high-definition videos from text or image prompts, while Imagen 3 creates detailed, lifelike images. Both models include customization and editing tools, designed to support applications, with safety measures such as digital watermarking and data governance built-in.

Veo, available in private preview, is Google DeepMind’s most advanced video generation model. It creates high-definition videos from text or image prompts, offering flexibility across a range of cinematic and visual styles. With a deep understanding of language and visual semantics, Veo produces videos that align with prompts while maintaining consistency and realism in scenes.

The model enables businesses to create video content more efficiently, reducing production time and costs. Google has demonstrated Veo’s ability to convert both real and AI-generated images into coherent video clips, offering practical applications for marketing, sales, and production teams to improve their visual storytelling.

Imagen 3, now generally available on Vertex AI, is Google’s highest-quality image generation model. It delivers lifelike images with remarkable detail and fewer visual artifacts compared to earlier versions. The model can generate high-definition images from simple text prompts, helping businesses create visuals tailored to their branding needs.

Imagen 3 also includes editing and customization features. Businesses can refine images with text prompts, adjust specific parts of images using mask-based editing, and upscale images to meet size requirements. Customization options enable users to infuse their brand identity into generated visuals, making it a valuable tool for marketing and design teams.

Google emphasizes that Veo and Imagen 3 were developed with safety and responsibility in mind, aligning with the company’s AI Principles. Both models include digital watermarking, safety filters to prevent harmful content, and robust data governance policies that ensure customer data is not used for training. 

The models are already being tested and adopted by companies across various sectors. Marketing leader WPP integrates Imagen 3 into its AI-powered WPP Open platform and plans to incorporate Veo for video content creation. Agoda is experimenting with combining Imagen and Veo for travel marketing campaigns.

The announcement of Veo and Imagen 3 has sparked excitement within the community. For example, one Reddit user shared: 

This is a huge step forward! With Google’s Veo and Imagen 3, the possibilities for video and image generation are expanding rapidly. Veo’s text-to-video capabilities are especially exciting—it’s incredible to think about how AI can now generate high-quality videos just from a simple script. I’ve been following these models closely, and it’s fascinating to see how tools like these are reshaping content creation.

To get started with Veo on Vertex AI, businesses can contact their Google Cloud account representative for access. More information is available in the documentation.