While Google’s Veo models have garnered international attention for their realistic video quality, its AI-powered video creation platform, Google Vids, has fallen by the wayside. A new slew of releases could change that.
On Thursday, Google launched a wave of AI-powered updates to Vids at no cost, with a focus on AI avatars. For instance, one of the biggest highlights is creating custom avatars from scratch, using prompts to describe what you want and then refining the appearance, including outfits, accessories and backgrounds.
These avatars can be created in seven languages: English, French, German, Italian, Korean, Portuguese, Spanish, and Japanese. Users can also more easily direct avatars to interact with objects or to place them in specific scenes. Lastly, new 2D and 3D cartoon-style avatars give users a fun alternative to human digital twins, which can sometimes feel like they're in the uncanny valley.
Ultimately, these customization features should help organizations more easily build avatars that match their brand identity, which is the most practical use case for AI avatars. There were also other, more subtle upgrades:
- General accessibility: All users with a personal Google account can now generate video clips in Vids via Veo 3.1. A single text prompt brings 8-second clips to life. Every account includes 10 video generations per month at no extra cost.
- YouTube: All Workspace customers and users with personal Google accounts can now export videos directly to YouTube from Vids.
- Chrome Extension: A new Google Vids Screen Recorder makes it easier to capture the screen, including audio or video from the browser.
- Music: Using Google’s AI music generation models, Lyria 3 and Lyria 3 Pro, users can now generate music for their projects as well.
This wave of updates follows Google’s launch of Veo 3.1 Lite, the company’s most cost-effective video model. This follows rapidly behind OpenAI’s decision to kill its AI video platform, Sora, though reports indicate the decision was made due to high compute costs. Of course, Google does not face the same compute constraints, since it owns one of the big three hyperscalers, Google Cloud.
Our Deeper View
OpenAI has pivoted away from video generation toward enterprise offerings in pursuit of greater profitability and strong market demand. Google's latest wave of updates, however, seems to occupy a middle ground. While Google Vids is fundamentally an AI video platform, its emphasis on AI avatars makes it far better suited for professionals and enterprise use cases. This positions it alongside companies like Synthesia and HeyGen, which have built entire businesses on a similar model, with Synthesia reporting $100 million in annual recurring revenue (ARR) as of April. If successful, this approach could quickly become the blueprint for monetizing AI video models while keeping profitability front and center.