Google launched its new video model Veo 3.1 with improved audio output, granular editing controls, and better output for image to video. It said that Veo 3.1 builds on May's Veo 3 release and generates more realistic clips and adheres to prompts better. The model allows users to add an object to the video and have it blend into the clip's style, Google said.
For example, OpenAI said in a blog post that the model was trained to be less overly optimistic, a characteristic that can be observed in instances where a Sora-generated video shows the player missing the shot but still making it into the hoop. With Sora 2, OpenAI claims the player would miss the shot, and the ball would rebound off the backboard.