Google has introduced an upgraded video model called Veo 3.1, featuring enhanced audio output, refined editing controls, and improved image-to-video conversion. According to the company, Veo 3.1 builds upon the version released in May and is capable of generating more realistic video segments with better responsiveness to user prompts.

The model enables users to insert objects into videos and seamlessly blend them with the visual style of the footage. Additionally, Google announced that an upcoming update to Flow will allow users to remove existing objects from videos.

Veo 3 already supports a range of editing tools, such as incorporating reference images to guide character animation, supplying start and end frames to generate AI-driven clips, and extending existing videos based on the final frames. With Veo 3.1, Google has added synchronized audio support to these features, making the output more engaging.

The company is now integrating this model into its video editor Flow, the Gemini app, as well as the Vertex and Gemini APIs. Google noted that since Flow's launch in May, users have created more than 275 million videos using the platform.