Google Vids Doubles Down on AI Video, Adding New Models and Directable Avatars

While some competitors scale back, Google is accelerating its push into AI-generated video. The company has integrated its most advanced video and audio models directly into Google Vids, its editing tool, and introduced a selection of customizable AI avatars for generated content. The update also streamlines sharing to YouTube.

The centerpiece is Veo 3.1, a model Google claims offers significant gains in visual realism and coherence. Although Veo has been marketed to professional creators, Google Vids is aimed squarely at everyday business and personal use—think promotional clips, event announcements, or personalized video messages. The tool remains free at a basic level, but generative features come with tight limits without a paid AI plan.

Free accounts are restricted to 10 AI video generations monthly. Subscribers to the AI Pro tier receive 50, while those on the premium AI Ultra plan, for individuals or companies, get 1,000. Output follows Veo's standard format: eight-second clips at 720p resolution.

Google has also baked its Lyria music models into Vids. The latest iteration generates short audio tracks based on a simple description of mood or style, no lyrics required. The result is functional, if not artistically profound, background music suitable for the tool's intended casual projects. Usage limits for AI music also increase with higher subscription tiers.

Source: Ars Technica