.jpg)
Generative artificial intelligence (AI) has been around for three years. At first, the generated results looked awkward and predictable, not real videos. Technology is advancing and the generated videos are becoming more realistic but have shortcomings because no audio is generated at all. At Google I/O, the new generative AI model Veo 3 was announced with the ability to generate videos with matching audio.
In the demo shown, audio that matches the generated video was successfully produced. Because of its quite good capabilities, Veo 3 is a model that can be accessed by Google Flow customers to generate short films starting today.
At this point, Veo 3 can only be accessed by Google AI Ultra customers for a fee of RM1229.90 per month. Some of the generated results that have been shared on social media look realistic but are not yet perfect.
Veo 3 is not the only generative AI model that can generate videos with matching audio. In October last year, Meta Movie Gen was also shown with the same capabilities but until now it has not been offered as a commercial product by Meta.