We’re introducing Veo, our most capable model for generating high-definition video, and Imagen 3, our highest quality text-to-image model. We’re also sharing new demo recordings created with our Music AI Sandbox.