Google has launched its new video model, Veo 3.1. This release features improved audio output, more granular editing controls, and better performance for image-to-video generation. The company stated that Veo 3.1 builds on the Veo 3 release from May, generating more realistic video clips and adhering to user prompts more accurately.
A key feature of the model allows users to add an object to a video and have it blend seamlessly into the clip’s existing style. Google also announced that a future update will enable users to remove existing objects from videos within its Flow editor.
Veo 3 already included several editing capabilities. These features allow users to add reference images to guide a character’s appearance, provide the first and last frame to generate a full clip with AI, and extend an existing video based on its final frames. With Veo 3.1, Google is enhancing all of these features by adding integrated audio to make the resulting clips more dynamic and lively.
The company is now rolling out the Veo 3.1 model to its video editor called Flow, the Gemini App, and through Vertex and Gemini APIs. Google reported that since Flow’s initial launch in May, users have already created more than 275 million videos on the platform.