Google launches Veo 3, a video model that generates audio, plus Imagen 4 and Flow
Veo 3 generates video with synchronised audio and dialogue, released to Gemini Ultra subscribers in the US alongside the Imagen 4 image model and the Flow tool.
Google has announced Veo 3, a video model capable of generating audio, together with the Imagen 4 image model and Flow, a tool for filmmaking. According to the announcement, Veo 3 is released to Gemini Ultra subscribers in the United States.
What is new in Veo 3 is generating video with synchronised audio and dialogue in a single pass, instead of building visuals and then adding voice and effects at separate stages. For corporate content creators, this shortens the distance from an idea to a finished clip. Imagen 4 handles the still-image side, while Flow organises the filmmaking workflow.
The practical significance lies in content production costs: a product introduction video or a training document once required a filming and editing crew, and can now start from a draft generated by the model. This also changes how work is divided within communications teams.
Enterprises should note three points: the release scope is still limited to one group of subscribers and one market; copyright and authenticity of images and audio generated by the model; and the content approval process before publication.
Source: blog.google
