Google Introduces Video, Image Generation AI Models Veo 2, Imagen 3

Google has introduced two cutting-edge AI models, Veo 2 for video generation and Imagen 3 for image generation, designed to deliver exceptional results. These models are now available through VideoFX, ImageFX, and the new Google Labs experiment, Whisk.

Veo 2 creates high-quality videos across a wide range of subjects and styles, achieving exceptional results in human-rated comparisons against leading models. This model enhances realism by understanding real-world physics and human movement, generating more detailed and lifelike videos. Users can request Veo 2 to produce specific cinematic effects, such as a low-angle tracking shot or a close-up of a scientist. It can generate videos in resolutions of up to 4K and durations lasting several minutes.

"While video models often “hallucinate” unwanted details -- extra fingers or unexpected objects, for example -- Veo 2 produces these less frequently, making outputs more realistic," Aäron van den Oord, Research Scientist, and Elias Roman, Senior Director, Product Management, Google Labs, write in the blog introducing Veo 2 and Imagen 3.

Imagen 3 is an upgraded AI image generation model, now available in ImageFX to over 100 countries. This latest iteration has received several upgrades, including the ability to render more diverse art styles, greater accuracy, and better-composed images.

To ensure responsible use, Google has added a native SynthID watermark on generated videos to help identify AI-generated content from real videos, reducing the risk of misuse for deepfakes. The company plans a gradual rollout of Veo 2 to control access and improve quality and safety before global availability.

Veo 2 is currently available to select users in VideoFX via Google Labs, with a waitlist for those interested in trying out the tool. Google plans to expand Veo 2 to YouTube Shorts and other products next year.