Sundar Pichai, CEO of Google, speaks on the Google I/O developer convention.
Andrej Sokolow | Image Alliance | Getty Photographs
Google on Tuesday introduced Veo 3, an AI video generator that may additionally create and incorporate audio.
The bogus intelligence device competes with OpenAI’s Sora video generator, however its capability to additionally incorporate audio into the video that it creates is a key distinction. The corporate stated Veo 3 can incorporate audio that features dialogue between characters in addition to animal sounds.
“Veo 3 excels from textual content and picture prompting to real-world physics and correct lip syncing,” Eli Collins, Google DeepMind product vice chairman, stated in a weblog Tuesday.
The video-audio AI device is offered Tuesday to U.S. subscribers of Google’s new $249.99 per thirty days Extremely subscription plan, which is geared towards hardcore AI lovers. Veo 3 can even be obtainable for customers of Google’s Vertex AI enterprise platform.
Google additionally introduced Imagen 4, its armed conflict image-generation device, which the corporate stated produces higher-quality photos by consumer prompts. Moreover, Google unveiled Stream, a brand new filmmaking device that permits customers to create cinematic movies by describing places, photographs and elegance preferences. Customers can entry the device by Gemini, Whisk, Vertex AI and Workspace.
The armed conflict launches come as imagery and video turn out to be well-liked use circumstances for generative AI prompts. OpenAI CEO Sam Altman in March said ChatGPT’s 4o picture generator was so well-liked that it brought about the corporate’s computing chips to “soften.” The corporate stated it needed to briefly restrict the function’s utilization.
Google has a combined monitor document relating to its AI picture mills. Final yr, the corporate needed to relaunch its Imagen 3 device after it surfaced traditionally inaccurate outcomes to customers’ prompts, inflicting widespread criticism. Co-founder Sergey Brin later said the mishap was resulting from a scarcity of “thorough testing.”
The Mountain View, California, firm additionally up to date its Veo 2 video generator to incorporate the flexibility for customers so as to add or take away objects from movies with textual content prompts. Moreover, Google opened its Lyria 2 music-generation mannequin to creators by its YouTube Shorts platform and companies utilizing Vertex AI.
