
At its annual Google I/O 2025 keynote, Google introduced Veo 3, a powerful new AI tool that can generate both video and audio.
The tool is designed to turn text or image prompts into realistic video clips that also include sound, something that OpenAI’s Sora, a rival in this space, doesn’t yet support.
Google Veo 3
According to Google, Veo 3 can create sounds, such as character dialogue, nature sounds, and even animal sounds, all synchronised with the video. The tool comes with lip syncing and physics-aware visuals, so there's a lot for creators and developers to like.
Veo 3 is now available for $249.99 a month on Google’s Ultra plan to users affluent enough to afford it in the US Business users can use Veo 3 on Google’s Vertex AI platform.
On the same day, Google announced Imagen 4, a new or upgraded image-generation tool and Flow, a new filmmaking assistant that users can use to create cinematic video, simply by telling Flow where the locations, shots, and styles would be.
Excitingly, these tools are being provided through platforms including Gemini, Vertex AI, Whisk and Workspace.
Google's move shows how visual and sound-based AI tools are becoming more popular. Earlier this year, OpenAI faced heavy demand for its ChatGPT-4o image feature, which pushed its hardware to the limit.
However, Google has had some stumbles too, last year it had to relaunch its Imagen 3 tool after it gave users wrong and controversial image results.